Identification of novel resorcinol amide derivatives as potent and

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 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

Journal of Medicinal Chemistry

Identification of novel resorcinol amide derivatives as potent and specific pyruvate dehydrogenase kinase (PDHK) inhibitors 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†, ‡,* †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 1These

Therapeutics Inc., 32 Songdogwahak-ro, Yeonsu-gu, Incheon 21984, Republic of Korea.

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 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

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.

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.


Page 2 of 61

Page 3 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


O

O Cl

F3C

N

OH

F3C HO

Cl

N

OH O

CN

O

Nov3r (2)

O

O

Radicicol (5)

NH2

O

O

O

Pfz3 (4)

Cl N

F

F

O

NH

S

O Cl

S

H N

Cl

AZD7545 (3)

N OH O

O N

O

DCA (1)

HO

Cl

H N

HO

O

N

S

N

S

O

S

N 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,



binding affinity was assessed by using a fluorescence polarization assay rather than by using an assay based on inhibition of kinase activity.21 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.

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.


Page 4 of 61

Page 5 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


Scheme 1. Synthesis of 14 and 15a-ga O

O

OMe

O

OMe

a

O

b

O

N

O

O

N

H N

HO O

N

OH O

13

R1

f

H N

HO

OAc O

12

OH

e

H N

AcO

N OAc O

11

OMe

d

NH2

OAc O

OAc O

9 O

AcO

N OH

10a

8

NO2

AcO

HN

NH2

OMe

c AcO

NO2

aReagents

OMe

O

OH O

14

15a-g

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

a

NH2

HN

16

17a-t AcO

OH

10b: 4-acetoxy 10c: 2-acetoxy

R2

b

R2

N

OAc O

OAc O

O

HO N

OH

10a

N

c AcO

OH

18a-t

N

R2

O

19a-t

O

O

N

c F AcO

N O

20a-b aReagents

O

b AcO

10b-c

N

O

N

F HO

N O

21a: 4-hydroxy 21b: 2-hydorxy

21a-b

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.



Structure and activity relationships 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. 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.


Page 6 of 61

Page 7 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


R1 H N

HO N OH

Entry

O OH

N H

aRadiometric

Entry

PDHK2

PDHK4

7.220

4.670

15d

1.960

14.750

15e

8.590

7.840

15f

9.910

12.260

15g

IC50 (µM) a

R1

O N

PDHK2

PDHK4

0.503

0.311

N

1.620

1.980

2.660

5.793

1.660

1.810

O

O

15b

14, 15a-g

O

O

15a

15c

IC50 (µM) a

R1

14

O

O

N H

N O

O

O N H

OH

N

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



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. 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


Page 8 of 61

Page 9 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


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.

Table 2. Inhibitory activities of 15d, 19a-t and 21a-b against PDHK1/2/3/4.



N

Cl N

O

F

F

O

NH

HO

O

19a F

19b

R2

HO

N

N

OH O

6

O

15d, 19a-t

PDHK IC50 (µM) a

H N

15d

O

F N

OH O

R2

N

HO

N

Entry

Page 10 of 61

Entry

PDHK IC50 (µM) a

R2

1

2

3

4

ND b

0.503

ND b

0.311

19j

0.109

0.206

0.186

0.066

19k

0.067*

0.113*

0.142*

0.052*

19l

O

21a: 4-hydroxy 21b: 2-hydorxy

O

1

2

3

4

0.140

0.184

0.858

0.775

0.048*

0.133*

0.067*

0.069*

0.020

0.252

0.255

0.052

0.024

0.193

0.181

0.082

0.036*

0.010*

0.059*

0.013*

0.014

0.098

0.005

0.011

0.042

0.018

0.090

0.008

0.057

0.095

0.098

0.089

0.024

0.132

0.090

0.087

0.020*

0.064*

0.064*

0.020*

0.009*

0.008*

0.080*

0.005*

0.042*

0.065*

0.026*

0.021*

Cl

21a

-

ND b

2.040

ND b

0.390

19m

O

21b

-

ND b

> 10

ND b

> 10

19n

O

ND b

0.126

ND b

0.169

19o

O

ND b

0.140*

ND b

0.273*

19p

S

ND b

0.110

ND b

0.099

19q

S

0.151

0.302

0.461

0.399

19r

S

0.113*

0.150*

0.136*

0.034*

19s

S

19h

ND b

0.373

ND b

0.234

19t

19i

ND b

0.772

ND b

5.330

6

Br

19c

19d

19e

19f

19g

Cl

CF3

NO2

CN

OH

I

Br

CN

Br N

Br

N

Br

aRadiometric

S

-

biochemical kinase assay. bND stands for Not Determined. *Average value of two

independent assays.


Page 11 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


Molecular dynamics simulation 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,Ndimethylbenzamide 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.



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.


Page 12 of 61

Page 13 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 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,Ndimethylbenzamide 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.



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.

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


Page 14 of 61

Page 15 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


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.

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



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. In vitro potency of resorcinol amide derivatives against cancer cells and reconstituted PDCreaction 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 p53deficient 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 offtarget 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


Page 16 of 61

Page 17 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


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 50

> 50

50

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 PDCreaction 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,


Page 19 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


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. Table 4. CYP inhibition profiles of 19g, 19n and 19t.

CYP IC50 (µM) Entry CYP1A2

CYP2C9

CYP2C19

CYP2D6

CYP3A4

19g

> 20

13.4

> 20

> 20

6.47

19n

> 20

> 20

4.3

> 20

1.4

19t

> 20

18.0

7.8

> 20

Identification of novel resorcinol amide derivatives as potent and

Recommend Documents