High-Potency Phenylquinoxalinone Cystic Fibrosis ... - ACS Publications

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

High-potency phenylquinoxalinone cystic fibrosis transmembrane conductance regulator (CFTR) activators Jung-Ho Son, Jie S. Zhu, Puay-Wah Phuan, Onur Cil, Andrew P. Teuthorn, Colton K. Ku, Sujin Lee, Alan S Verkman, and Mark J. Kurth J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01759 • Publication Date (Web): 23 Feb 2017 Downloaded from http://pubs.acs.org on February 24, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

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

High-potency phenylquinoxalinone cystic fibrosis transmembrane conductance regulator (CFTR) activators Jung-Ho Son†, Jie S. Zhu†, Puay-Wah Phuan‡, Onur Cil‡, Andrew P. Teuthorn†, Colton K. Ku†, Sujin Lee‡, Alan S. Verkman‡, Mark J. Kurth†* †

Department of Chemistry, University of California, Davis, CA, 95616, USA

‡

Departments of Medicine & Physiology, University of California San Francisco, CA 94143, USA

ABSTRACT: We previously identified phenylquinoxalinone CFTRact-J027 (4) as a cystic fibrosis transmembrane conductance regulator (CFTR) activator with an EC50 of ~200 nM, and demonstrated its therapeutic efficacy in mouse models of constipation. Here, structure-activity studies were done on 36 synthesized phenylquinoxalinone analogs to identify compounds with improved potency and altered metabolic stability. Synthesis of the phenylquinoxalinone core was generally accomplished by condensation of 1,2-phenylenediamines with substituted phenyloxoacetates. Structure-activity studies established, among other features, the privileged nature of a properly positioned nitro moiety on the 3-aryl group. Synthesized analogs showed improved CFTR activation potency compared to 4 with EC50 down to 21 nM and with greater metabolic stability. CFTR activators have potential therapeutic indications in constipation, dry eye, cholestatic liver diseases, and inflammatory lung disorders.

ACS Paragon Plus Environment


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

Page 2 of 37

INTRODUCTION The cystic fibrosis transmembrane conductance regulator (CFTR) is a cAMP-regulated chloride channel expressed in mammalian epithelia in the respiratory, gastrointestinal, and reproductive systems, as well as in exocrine glands and other tissues.1 Loss-of-function mutations in CFTR cause cystic fibrosis, and CFTR over activation causes certain secretory diarrheas including cholera and Travelers’ diarrhea.2 CFTR is considered an important drug target, with activators of CFTR of potential benefit for constipation,3,4 dry eye,5 inflammatory lung disorders,6 and cholestatic liver disease; inhibitors of wildtype CFTR may be useful for treatment of certain secretory diarrheas and polycystic kidney disease;7,8 and correctors and potentiators of mutant CFTRs for treatment of cystic fibrosis.9 We previously identified by high-throughput screening the phenylquinoxalinone CFTRactJ027 (4; Figure 1) as a CFTR activator and demonstrated its efficacy in normalizing stool output, hydration, and intestinal transit in a mouse model of opioid-induced constipation.3 Phenylquinoxalinone 4 activated CFTR chloride conductance with an EC50 of ~200 nM and showed no apparent off-target actions or toxicity following chronic administration in mice. In a follow-up study,4 4 was shown by patch-clamp and biochemical studies to target CFTR directly, and was demonstrated to activate CFTR in human enterocytes and normalize stool parameters in mouse models of acute and chronic constipation. Side-by-side comparisons of intestinal fluid secretion and

H 2N N N

NO 2 O

CFTRact-J027 (4) EC 50 = 200 nM

Figure 1. A phenylquinoxalinone CFTR activator identified by high-throughput screening.


Page 3 of 37

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


stool output in constipation models showed greater efficacy of 4 than supramaximal doses of the FDA-approved drugs lubiprostone and linaclotide. Here, motivated by the potential therapeutic utility of phenylquinoxalinone-based CFTR activators in constipation and other diseases, we synthesized 36 analogs of phenylquinoxalinone 4 in order to establish structure-activity relationships and to identify compounds with greater potency. Also, while the rapid hepatic metabolism of 4 results in minimal systemic exposure following oral administration in mice, which is desirable for treatment of constipation, we also sought phenylquinoxalinone CFTR activators with greater metabolic stability for treatment of lung and liver disorders where systemic exposure is necessary.

RESULTLS AND DISCUSSION Chemistry. General synthesis of phenylquinoxalinones. Most of the phenylquinoxalinone in this study were expediently synthesized in four steps starting from acetophenones (Scheme 1). We generated the phenylquinoxalinone core by condensing o-phenylenediamines with substituted phenyloxoacetates (6), which were synthesized following literature methods.10 Scheme 1. Synthesis of phenylquinoxalinones 1-3. R1

O

a

R1

O Br

b

Br

R1

O MeO

O

5

6 NH 2 R2

c NH 2

R1

N

d

R2 N R3 1-3

O

R1

N R2 N H

O

7

(a) Br2, 1,4-dioxane; (b) DMSO, D; MeOH; (c) toluene, D; (d) R-Br, K2CO3, DMF.



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

Page 4 of 37

Briefly, substituted acetophenone was doubly brominated with bromine in 1,4-dioxane to give 5, then heated in DMSO followed by addition of methanol to give 6. Phenylquinoxalinone 7 was N1-alkylated using K2CO3 and R-X in DMF11 and pure products (1-3) were obtained via column chromatography. Phenylquinoxalinone 1a was prepared as outlined in Scheme 2. Treatment of methyl 3fluoro-4-nitrobenzoate (8) with methyl 2-cyanoacetate under basic conditions delivered intermediate 9.12 Subsequent copper(I) iodide-catalyzed aerobic oxidation12 delivered methyl 2-oxo-2phenylacetate 10 and, from here, target 1a was prepared in parallel to the chemistry employed in Scheme 1. With 1a in hand, saponification and nitro reduction were accomplished as outlined in Scheme 3 to deliver analogs 1b, 1j, and 1k.

Scheme 2. Synthesis of phenylquinoxalinone 1a.

O2N

O2N

a

NC

CO2Me NC

F

CO2Me

8

b

O2N O

CO2Me CO2Me

CO2Me

MeO

9

O

10

NH2

c

R2 NH2

O2N

O2N

N

CO2Me

R2

N

O

1a (R2 = H)

d

N

CO2Me

R2

N H

O

11

(a) Cs2CO3, DMSO; (b) CuI, 1,10-phen, ACN, O2; (c) toluene, D; (d) K2CO3, DMF, BnBr.


Page 5 of 37

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 3. Diversification from phenylquinoxalinone 1a. O2N

H 2N

N N

b

N

COOR O

1a (R = Me) 1b (R = H)

COOR

a N

c

O

1j (R = Me) 1k (R= H)

(a) H2, Pd/C, MeOH; (b) NaOH, EtOH; (c) KOH, MeOH. Constrained ring phenylquinoxalinone analogs 16 and 20 were prepared as outlined in Scheme 4. 2-Amino-3-nitrophenol was N- and O-alkylated with 1,2-dibromoethane to deliver intermediate 14 and subsequent nitro reduction and condensation of the resulting diamine with 1-acetyl-5-nitroindoline-2,3-dione led smoothly to analog 16.13 Employing 2-bromo-1-phenylethan-1-one in place of 1,2-dibromoethane and 5-nitroindoline-2,3-dione in place of 1-acetyl-5nitroindoline-2,3-dione delivered analog 20.14 Interestingly, the reaction of N-benzyl-1,2-diaminobenzene with 5-fluoroisation (in analogy with the protocol employed to prepare compounds 16, 20 and 4) led to 22 and the attempted de-acylation of 21 (X = NO2) led to 23.



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

Page 6 of 37

Scheme 4. Synthesis of constrained ring phenylquinoxalinone analogs. NO2

12 OH

NH2

NO2

NH2

a

b

NH

NH

Br

O

O

Br

13

14

15

O

c

NO2 NH2 OH

d

NH

Ac N

Br

O

O

H N NO2

N

18

NO2

O

12 NO2

Ac N

O

O

17

16

b

O

NH2

H 2N

H N

N NO2

O

NH

N

19

F3C

O

O

c

O

NO2

20

O

NH2

H N

F F

O

NH

F3C

N N

N

c

22

F3C

Ac N N

21

H N

NO2 NO2 O

F3C

N

e N

N

23

(a) KOH, DMF; (b) H2, Pd/C, MeOH; (c) HOAc, toluene; (d) K2CO3, ACN (e) HCl, MeOH. Modifications of the 3-aryl ring. Compound 4 contains a 2-amino-5-nitro phenyl ring at the 3-position of the quinoxalinone core (Figure 1) and our first effort was to modify this ring. Several compounds (Table 1) were rationally synthesized and their activities were determined using a plate reader assay. The most active compound was CFTRact-J125 (1c), which only lacks the 2-


Page 7 of 37

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


amino group at C2 of the 3-aryl ring compared to 4, and it showed an ~10-fold increased potency when compared to 4. We then synthesized a series of analogs retaining this amino group deletion. In contrast to the high activity of 1c, compounds without the 3-nitro group had significantly lower activity (1h), indicating the privileged nature of the 3-nitro group in 1c. Table 1. CFTR activation with variation in the 3-aryl ring (EC50 reported in µM ± S.E.M.). R1

N 3

N

O2N

O

O2N CO2Me

NO2

CO2H

1a EC50 = 9.9 ± 2

1b EC50 = >10

O2N

1c (J125) EC50 = 0.009 ± 0.003

CF3

NO2

1d EC50 = 1 ± 0.2

1e EC50 = 1 ± 0.1

1f EC50 = 1 ± 0.2 F

Br

NO2

1g EC50 = 0.33 ± 0.07

1h EC50 = 5 ± 0.4

1i EC50 = 0.69 ± 0.22

Nitro bioisosteres H 2N

H 2N CO2Me

1j EC50 = 0.55 ± 0.1

CO2H

1k EC50 = >10 N O

CN

1l EC50 = 1 ± 0.4

N

1m EC50 = 0.018 ± 0.006

Bioisosteric compounds containing –COOR replacements of the –NO2 moiety in the 3-aryl group were synthesized according to Schemes 2 and 3. It was found that carboxylic acid analog



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

Page 8 of 37

1k was inactive, while methyl ester analog 1j was moderately potent. Combining a 2-NO2 with either a 5-COOMe or a 5-COOH (1a and 1b, respectively) resulted in very low activity. Positional deviation of the nitro group also resulted in reduced activity, as with 2-nitro or 4-nitro analogs 1d and 1e, respectively. Introduction of other functional groups, such as 4-CF3 or 3-Br, in place of the 3-NO2 of 1c also showed low activity (1f and 1g, respectively). Introducing a fluorine atom into 1c (i.e., 1i) reduced activity. Replacements of the nitro group with a bioisosteric nitrile (1l) greatly reduced activity, but interestingly, the bioisosteric benzoxadiazole (1m) showed comparable potency to the nitro analog (1c). Modifications of the quinoxalinone core. Moving forward with 1c as the lead, quinoxalinone backbone modifications were undertaken (Table 2). Halogen substitution at the 6-position, such as 6-Cl (2b) or 6-Br (2f), showed good activity albeit less than that of 1c. Substitution at the 5-, 7-, or 8-positions generally resulted in lower activity. Disubstitution at the 6- and 7-positions (dichloro; 2d / difluoro; 2g / dimethyl; 2i) reduced activity. Table 2. CFTR activation with variation in the quinoxalinone core (EC50 reported in µM ± S.E.M.). N

NO2

R2 N

O

Cl Cl Cl

2a EC50 = 0.76 ± 0.38

2b EC50 = 0.07 ± 0.03

Br

Cl Cl

2d EC50 = >10

2c EC50 = 0.11 ± 0.04

Br

2e EC50 = Inactive

2f EC50 = 0.063 ± 0.03

2h EC50 = >10

2i EC50 = >10

F F

2g EC50 = >10


Page 9 of 37

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


Modifications of the N1-substituent. We next modified the N1-benzyl group (Table 3). We reported previously that placing substituents on the benzyl ring was not tolerated.4 However, heterocycle analogs such as thiophenyl (3j), furanyl (3i), and pyridyl (3h) showed comparable activity. Among short chains, allyl and ethyl groups showed moderate activity, but methyl and propyl groups had poor activity (3a-e). Replacing the phenyl with a larger aromatic group, such as naphthyl, significantly reduced activity (3f and 3g).

Table 3. CFTR activation with variation in the N1-substituent (EC50 reported in µM ± S.E.M.). N N

NO2 O

R3

H 3a EC50 = >10

CH3 3b EC50 = >10

3e EC50 = 0.055 ± 0.01 N

3h EC50 = 0.052 ± 0.02

3c EC50 = 0.065 ± 0.02

3f EC50 = Inactive

O

3i EC50 = 0.02 ± 0.01

3d EC50 = 0.177 ± 0.048

3g EC50 = 4.3 ± 0.6

S

3j (J170) EC50 = 0.01 ± 0.005

Compounds with constrained rings. Motivated by our prior findings that constrained rings can enhance the activity of a CFTR corrector,15 we next examined two different types of constrained analogs (Scheme 4) with hindered rotation of the N1-alkyl group (16 and 20) or phenyl ring (22 and 23). As previously shown, the N-acetylated version of 4 has very low activity;4 we therefore attempted to synthesize 16 without an N-acyl group, but both deacylation of 16 and



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

Page 10 of 37

condensation with 5-nitroisatin failed. The activities of neither 20 (EC50 = >10 µM) nor 16 (EC50 = 0.82 µM) were greater than that of 4 (see Table 4). Phenyl-ring constrained compounds 22 and 23 were unexpected by-products of the deacylation reaction (see Scheme 4 and Table 4). These compounds are purple and red, respectively – a consequence of their extended aromatic systems – and those colors are different from the bright yellow color of most derivatives of 4. Intramolecular heterocycle formation in this system might have been facilitated by the presence of the electron withdrawing CF3 group in the quinoxalinone backbone under these acidic deacylation conditions. With other substituents in the quinoxalinone backbone, only small amounts of uncharacterized reddish byproduct formed during the deacylation step, suggesting a minimal amount of by-product formation.

Table 4. CFTR activation with constrained ring analogs (EC50 reported in µM ± S.E.M.). H 2N N N

Ac N

NO2 O

O

N

H N NO2 O

O

20 EC50 = >10

16 EC50 = 0.82 ± 0.25 X

F3C

N N

N

X = F; 22 EC50 = >10 X = NO2; 23 EC50 = >10

Biology. In vitro characterization of phenylquinoxalinones. Phenylquinoxalinone CFTR activators with the highest potency as determined by plate-reader assay, 1c and CFTRact-J170 (3j; Table 3), were further characterized. Short-circuit current measurements were done using CFTR-expressing FRT cells in the presence of a transepithelial chloride gradient and with permeabilization of the cell basolateral membrane; consequently, current is a direct, linear measure


Page 11 of 37

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


of CFTR chloride conductance. Compounds 1c and 3j were added to the apical side of the monolayer. Representative data in Figure 2 for 1c and 3j shows a small increase in current following addition of a low concentration of forskolin, followed by concentration-dependent increases in current following activator additions. EC50 values were determined to be 21 ± 4 and 70 ± 11 nM for 1c and 3j, respectively. The potency for CFTR inhibition by 1c was similar in nonpermeabilized and permeabilized CFTR-expressing FRT cells (data not shown).

Figure 2. Short-circuit current measurement of CFTR activation by 1c and 3j. A. Measurements done in FRT cells expressing human wild-type CFTR showing responses to indicated concentrations of forskolin, 1c or 3j, and 10 µM CFTR inhibitor CFTRinh-172. B. Concentration-dependent activation of CFTR (mean ± S.E.M., n=3).

The CFTR specificity of the most potent compound, 1c, was further studied. At 10 µM, 1c did not affect the cellular cAMP level (Figure 3A), nor did it elevate cytoplasmic calcium or inhibit the ATP-stimulated elevation in cytoplasmic calcium (Figure 3B). In addition, 1c at 10 µM neither inhibited nor activated calcium-activated chloride channels in HT-29 cells (Figure 3C) or in FRT cells expressing TMEM16A (Figure 3D).



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

Page 12 of 37

Figure 3. In vitro characterization of 1c. A. Cellular cAMP in FRT cells in response to incubation for 10 min with 10 µM 1c without or with 90 nM forskolin (fsk). Positive controls included fsk (100 nM and 20 µM) and fsk + IBMX (20 µM + 100 µM) (mean ± S.E.M., n = 4). B. Cytoplasmic calcium measured by Fluo-4 fluorescence. FRT cells were pretreated for 5 min with 10 µM 1c (or control), with 100 µM ATP added as a calcium agonist as indicated. C. CaCC activity measured in HT-29 cells expressing YFP showing no activation (iodide addition) or inhibition (iodide + ATP addition) by 10 µM 1c. D. TMEM16A activity measured in FRT cells expressing YFP showing no activation (iodide addition) or inhibition (iodide + ATP addition) by 10 µM 1c.

Efficacy of 1c in a loperamide-induced mouse model of acute constipation. We previously demonstrated the efficacy of 4 in a loperamide-induced mouse model of constipation.3,4 Here, the efficacy of 1c was tested. Phenylquinoxalinone 1c was administered orally to mice 1 h prior to loperamide and 3-h stool samples were collected after loperamide. Figure 4 shows that orally administered 1c fully normalized stool weight, pellet number and hydration with halfmaximal effective dose (ED50) < 1 mg/kg.


Page 13 of 37

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 4. Efficacy of 1c in a mouse model of acute constipation. A. Experimental protocol. Mice were treated with 1c (orally) or vehicle 1 hour before loperamide (0.3 mg/kg, intraperitoneal). Weight, pellet number and water content were determined for stool collected at 3 h. B-D. Stool weight, pellet number and water content after treatment with 0, 0.1, 0.3, 1, 3, 10 mg/kg of 1c (mean ± S.E.M., 4 mice per group). The control was without treatment with loperamide and 1c.

Phenylquinoxalinone 4 was shown previously to have minimal oral bioavailability and rapid metabolism.3,4 Though these properties are favorable for ‘topical’ applications in the treatment of constipation and dry eye in which systemic exposure is not needed, they are not favorable for treatment of liver and lung diseases where systemic exposure and organ accumulation are desired. Figure 5 shows substantially slower in vitro hepatic microsomal metabolism of 1c compared with 4, with nearly 80% of 4 metabolized at 60 min compared with

High-Potency Phenylquinoxalinone Cystic Fibrosis ... - ACS Publications

Recommend Documents