Page 1 of 82 ACS Paragon Plus Environment ... - ACS Publications

Subscriber access provided by Washington University | Libraries

Article

Discovery of Potent and Centrally Active 6-Substituted 5-Fluoro-1,3-dihydrooxazine #-Secretase (BACE1) Inhibitors via Active Conformation Stabilization Kenji Nakahara, Kouki Fuchino, Kazuo Komano, Naoya Asada, Genta Tadano, Tsuyoshi Hasegawa, Takahiko Yamamoto, Yusuke Sako, Masayoshi Ogawa, Chie Unemura, Motoko Hosono, Hisanori Ito, Gaku Sakaguchi, Shigeru Ando, Shuichi Ohnishi, Yasuto Kido, Tamio Fukushima, Deborah Dhuyvetter, Herman Borghys, Harrie J.M. Gijsen, Yoshinori Yamano, Yasuyoshi Iso, and Ken-ichi Kusakabe J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00011 • Publication Date (Web): 18 May 2018 Downloaded from http://pubs.acs.org on May 18, 2018

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 82 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

Yamano, Yoshinori; Shionogi Pharmaceutical Research Center, Discovery Research Laboratory for Core Therapeutic Areas Iso, Yasuyoshi; Shionogi Pharmaceutical Research Center, Discovery Research Laboratory for Core Therapeutic Areas Kusakabe, Ken-ichi; Shionogi & Co., Ltd., Medicinal Chemistry

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

Page 2 of 82

Discovery of Potent and Centrally Active 6Substituted 5-Fluoro-1,3-dihydro-oxazine βSecretase (BACE1) Inhibitors via Active Conformation Stabilization Kenji Nakahara,†,# Kouki Fuchino,†,# Kazuo Komano,† Naoya Asada,† Genta Tadano,† Tsuyoshi Hasegawa,† Takahiko Yamamoto,† Yusuke Sako,† Masayoshi Ogawa,† Chie Unemura,† Motoko Hosono,† Hisanori Ito,† Gaku Sakaguchi,† Shigeru Ando,‡ Shuichi Ohnishi,‡ Yasuto Kido,‡ Tamio Fukushima,‡ Deborah Dhuyvetter,║ Herman Borghys,║ Harrie J. M. Gijsen,§ Yoshinori Yamano,† Yasuyoshi Iso,† and Ken-ichi Kusakabe*,† †

Discovery Research Laboratory for Core Therapeutic Areas, ‡Research Laboratory for

Development, Shionogi Pharmaceutical Research Center, 1-1 Futaba-cho 3-chome, Toyonaka, Osaka 561-0825, Japan §

Neuroscience Medicinal Chemistry, ║Discovery Sciences, Janssen Research & Development,

Turnhoutseweg 30, B-2340 Beerse, Belgium


1

Page 3 of 82 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


ABSTRACT

β-Secretase (BACE1) has an essential role in the production of amyloid β peptides that accumulate in patients with Alzheimer’s disease (AD). Thus, inhibition of BACE1 is considered to be a disease-modifying approach for the treatment of AD. Our hit-to-lead efforts led to a cellular potent 1,3-dihydro-oxazine 6, which however inhibited hERG and showed high P-gp efflux. The close analog of 5-fluoro-oxazine 8 reduced P-gp efflux; further introduction of electron withdrawing groups at the 6-position improved potency and also mitigated P-gp efflux and hERG inhibition. Changing to a pyrazine followed by optimization of substituents on both the oxazine and the pyrazine culminated in 24 with robust Aβ reduction in vivo at low doses as well as reduced CYP2D6 inhibition. Based on the X-ray analysis and the QM calculation of given dihydro-oxazines, we reasoned that the substituents at the 6-position as well as the 5fluorine on the oxazine would stabilize a bioactive conformation to increase potency.

INTRODUCTION Alzheimer’s disease (AD) is the most common cause of dementia, accounting for approximately 60% to 50% of all dementia cases.1 In 2015, 46.8 million people worldwide were suffering from dementia.2 According to a survey in the United States, AD is the sixth leading cause of death, with 11% of people aged 65 and older suffering from AD. Total costs in 2016 for patients with AD and other dementias were estimated at $236 billion in the U.S., of which the national care service is expected to cover 68%. AD ultimately leads to death within 3−9 years after diagnosis. More significantly, there is a great impact on the quality of life for patients as well as their caregivers. Indeed, a study indicated that 40% of family caregivers suffer from depression, and as a result, 72% of them experienced relief when the dementia patients died.1 Unfortunately,


2


Page 4 of 82

none of the medications available today for AD slows or stops the disease progression, which points to a great unmet medical need for disease-modifying therapy for AD. The accumulation of β-amyloid plaques, which are composed of amyloid β (Aβ) peptides of 38−43 amino acids, and an abnormal form of hyper-phosphorylated tau proteins are the histopathological hallmarks of AD.1,3 Therefore, targeting inhibition of Aβ peptide production is considered to be a key strategy for disease-modification of AD.4 The β-site APP cleaving enzyme (BACE1), also known as β-secretase or memapsin2, is a rate-limiting enzyme in Aβ peptide production.5 BACE1, an aspartic protease, cleaves the amyloid precursor protein (APP) to generate a soluble N-terminal domain of APP (sAPPβ) followed by cleavage of the membrane-bound C-terminus of C99 by γ-secretase to generate the Aβ peptides. Among all isoforms of Aβ (total Aβ), the Aβ42 consisting of 42 amino acids is the most toxic form, as it readily aggregates to generate toxic oligomers which ultimately result in deposition of amyloid plaques.6 Although γ-secretase is a potential target for therapeutics to reduce the Aβ peptide, inhibition of the processing of other substrates, such as Notch, raises concerns about safety in clinical use, and its modulation suffers from lack of efficacy.7 Thus, reducing Aβ by inhibiting BACE1 is considered to be an attractive approach to disease modification. Genetic evidence also supports the importance of BACE1 inhibition in AD. The Swedish mutations in APP of K670N and M671L, close to the cleavage site in BACE1, cause familial AD by increasing the processing and production of total Aβ.6 The increased production of APP as observed in people with Down syndrome also results in early-onset AD, due to the rapid accumulation of Aβ.1 Furthermore, an APP variant of A673T has been found to be protective against AD, as it reduces the APP processing by BACE1 resulting in approximately 40%


3

Page 5 of 82 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


inhibition of Aβ production.8 Such evidence provides further support for BACE1 as a promising AD target. Since the discovery of BACE1 in 1999, a number of pharmaceutical companies and academic institutions have focused on exploring potent and brain penetrant BACE1 inhibitors.9−11 Because of the challenging nature of the target, the large size of the catalytic site, attention was initially directed at peptidomimetic inhibitors, such as statine- and hydroxyethylamine-based analogs derived from its transition state. Though the peptidomimetics were found to be potent in biochemical or binding assays, their low cellular potency as well as high P-gp efflux resulted in poor central efficacy in relevant preclinical models.9 After the identification of amidine-based small molecule inhibitors,12 many research groups have focused on the optimization of this class of BACE inhibitors.11 These efforts have overcome the issues observed with the peptidomimetics and led to the development of clinical BACE1 inhibitors, such as LY-2811376,13 LY-2886721,14 verubecestat (1, MK-8931),15 elenbecestat (2, E-2609),16 RG-7129,17 AZD-3839,18 lanabecestat (AZ-3293),19a CNP520,19b and JNJ-54861911.20 Some of the clinical inhibitors achieved robust Aβ reduction in humans, while compounds such as LY-2811376, LY-2886721, and AZD-3839 failed due to off-target related toxicity,13,14 indicating the need for other chemotypes to ensure successful development of BACE1 inhibitors for the market.


4


Page 6 of 82

Chart 1. Representative BACE1 Inhibitors

F

H2N N

H N O

N N

O S O

F

H2 N N

F

H N

N F

1 (verubecestat)

O

S H

N O F

2 (elenbecestat)

In contrast to fragment-based drug discovery programs pursued by other companies,13,15 our early research efforts began with a successful high throughput screen campaign of our corporate compounds utilizing a cellular Aβ production assay. This delivered the hit compound 3,21 which was contingently the same as that of Roche’s;22 their efforts culminated in compound 4 with significant CSF Aβ reduction in rat at a low dose. Our subsequent hit-to-lead SAR effort yielded initial leads 5 and 6. Because these leads had distinct profiles in terms of potency, P-gp recognition, and brain penetration, our strategy to identify BACE1 clinical candidates was to optimize the two leads in parallel; work on the thiazine 5 will be reported elsewhere. Dihydrooxazine 6 demonstrated good in vitro potency while displaying suboptimal hERG selectivity and high P-gp efflux, which resulted in low brain penetration and no significant efficacy in our mouse model. Lowering pKa has been proven to be an effective approach to reduce both hERG inhibitory activity and P-gp efflux.23 In another paper, we describe how we adopted this strategy to address these issues observed in 6 by inserting an olefin bond and following this with further optimization. This successfully reduced both the pKa and P-gp efflux leading to oxazine 7, which displayed robust Aβ reduction in both mouse and dog at low doses.24 Herein, we describe an alternative approach to mitigate the drawbacks of 6 by substituting the dehydro-oxazine head group with electron withdrawing groups followed by optimization of the tail group. This


5

Page 7 of 82 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


ultimately led to compounds 21 and 24 that showed robust Aβ reduction in vivo at low doses with reduced hERG and CYP2D6 inhibition. Chart 2. HTS Hit Compound 3, Roche Compound 4, Early Lead Compounds 5 and 6, and Oxazine 7

RESULTS AND DISCUSSION Exploration at the 6-Position of Dihydro-Oxazine 8 Leading to 13. Dihydro-oxazine 6 exhibited IC50 values of 118 nM in the biochemical assay and 3.7 nM in the cellular assay. The ability of 6 to reduce total Aβ levels in the brain was examined with wild-type mouse using po dosing at 10 mg/kg, where no significant Aβ reduction (15%) was observed at 6 h with a brain concentration of 29 ng/mL and a brain-to-plasma (B/P) ratio of 0.22. The poor efficacy and brain penetration were rationalized by its high P-gp efflux ratio of 20, which we postulated could be avoided by lowering the high basicity of the amino-dihydro-oxazine (pKa = 9.8). Another drawback of 6 was a high hERG inhibitory activity of 66% at 3 µM in an automated patch-clamp assay.


6


Page 8 of 82

During the course of our research, the Roche group reported that they had successfully lowered the pKa of 5-fluorinated oxazine 822b,25 to 8.1 as well as having increased its biochemical potency by 2-fold relative to 6 (Table 1). Importantly, the reduced pKa observed for 8 was associated with a mitigated P-gp efflux ratio of 4.0. Although 8 was still a P-gp substrate and showed a comparable hERG potency relative to 6, we considered 8 as a good template for further exploration. According to analysis using our proprietary BACE1 inhibitors, the lower limit of pKa for retaining cellular activity was 6.5.24 Therefore, we sought to further decrease the basicity of 8 by introducing electron withdrawing groups at the 6-position on the dihydro-oxazine as shown in Table 1. Prior to the investigation, we explored a small set of substituents at the 5position of 8. Consistent with Roche’s SAR, the compound with the opposite stereochemistry (10) exhibited pronounced reduction in potency.22b Because the fluorine at the 5-position showed good van der Waals interaction with Tyr132,22b a small substituent such as a methyl group was also expected to improve potency. Though 9 was less active than the corresponding F-analog 8, a comparable potency was unexpectedly observed for 11 regardless of its opposite configuration at the 5-position to 9. We then pursued electron withdrawing groups at the 6-position in 8. As outlined in Table 1, introduction of the fluoromethyl group (13) imparted a 5-fold increase in activity, and a reduced pKa of 7.3 translated into a lower P-gp efflux and hERG inhibition. The corresponding diastereomer at the 6-position (12) caused a 10-fold decrease in activity, while the P-gp efflux and hERG activity were slightly mitigated probably due to the reduction in the pKa by 0.2 unit. Encouraged by the improved profile observed for 13, we incorporated an alternative of the fluorine such as methoxy, which led to compounds 14 and 15. Unfortunately, these changes were accompanied by a moderate loss of biochemical potency as well as increased P-gp recognition.


7

Page 9 of 82 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


The increased P-gp efflux can be rationalized by the concomitant penalty of the higher basicity and the increased hydrogen bonding potential by the methoxy group relative to 13. Taken together, compound 13 with a low pKa of 7.3 achieved an optimal balance of potency, Pgp efflux, and hERG inhibition. Like 13, the lowest values of P-gp efflux and hERG inhibition were observed in 12 with the lowest pKa value in Table 1, though its potency was reduced. However, with the other compounds having pKa values of 7.6 or higher, the lower pKa values (lower basicity) did not necessarily translate into reduced P-gp efflux and hERG inhibition. For example, compounds 10 and 15 with the same pKa value of 7.6 strongly inhibited hERG relative to the other compounds with higher pKa values (higher basicity). In addition, a significant difference in P-gp efflux was observed between 10 and 15, despite their having the same pKa value. Matched pair analysis of 5-fluoro-oxazines 8 and 10 indicated that 10 possessed higher hERG inhibitory activity than more the basic 8. These results imply that subtle structural modification might be a mitigation strategy for hERG and P-gp efflux for basic compounds like those shown in Table 1 with higher pKa values at the physiological pH (7.4), because such basic compounds are largely ionized at the physiological pH (e.g., 80% of compounds with a pKa of 8.0 are ionized at pH 7.4). To assess the pharmacokinetic (PK) profiles including brain penetration for the representative compounds in Table 1 of 6, 13 and 14, we utilized our screening rat PK studies using a cassette dosing method (Table 2). Prior to the investigation, the metabolic stabilities in rat microsomes for these compounds were determined to be in an acceptable range. The compounds were dosed intravenously (iv) at 0.5 mg/kg and orally (po) at 1 mg/kg. The brain concentration was measured at the 0.5 h time point in the iv study. As expected, compound 13, with an improved Pgp efflux relative to 6 and 14, achieved an increased B/P ratio of 0.62 as compared to 6 and 14


8


Page 10 of 82

(B/P = 0.29 and 0.17, respectively). Moreover, a favorable PK profile was demonstrated in 13, with a low total clearance (CL) of 5.3 mL/min/kg and a high oral bioavailability (F) of 73% resulting in a decent maximal plasma concentration (Cmax) of 101 ng/mL at 1 mg/kg. In parallel with the PK study, the cellular IC50 of 13 was determined to be 0.58 nM. Although compound 13 was demonstrated to be potent and brain penetrable, its hERG inhibitory activity was still in a suboptimal range (55% inhibition at 3 µM).


9

Page 11 of 82 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


Table 1. Exploration of the Head Group

a

Values represent the mean values of at least two determinations. Biochemical homogeneous time-resolved fluorescence (HTRF)-based assay. The IC50 value for the reference compound (cas# 797035-11-1) was 22.6 nM. bEfflux ratio measured in LLC-PK1 cells transfected with human MDR1. c% inhibition at 3 µM measured in CHO cells transfected with hERG channels using an automated patch clamp system. dpKa determined by capillary electrophoresis.


10


Page 12 of 82

Table 2. Pharmacokinetic Properties of Key Compounds for Sprague-Dawley Rat

compd

RLM (%)c

fu,s

d

P-gp ERe

rat, iv, 0.5 mg/kg, n = 2a

rat, po, 1 mg/kg, n = 2b

CL (ml/min/kg)f

Vdss (L/kg)g

B/Ph

AUC (ngh/ml)i Cmax (ng/ml)j

F (%)k

37

6.1

0.29

328

32

73

4.5

0.62

2390

101

73

4.4

0.17

166

29

40

6

104

0.41

13

97

0.27

3.0

14

78

0.41

6.3

20

107

0.33

1.4

8.3

3.5

1.6

885

75

44

21

101

ND

1.4

9.3

5.4

1.9

660

58

35

22

100

ND

0.79

6.7

5.4

1.4

992

64

39

24

105

ND

1.8

6.9

2.2

630

50

43

20

5.3 41

11

a

Dosed as a solution of test compounds in N,N-dimethylacetamide (DMA)/propylene glycol (PG) = 1/1. bDosed as a suspension of test compounds in 0.5% methylcellulose (MC, 400 cP). c% remaining in rat liver microsomes after 30 min incubation. dFraction unbound in rat serum. ND = not determined; the fu values for 21, 22, and 24 were not determined due to their instability in rat plasma. eEfflux ratio measured in LLC-PK1 cells transfected with human MDR1. fTotal clearance. gVolume of distribution at steady state. hTotal brain-to-plasma ratio. iPlasma area under the curve. jMaximal plasma concentration. kOral bioavailability.

Optimization of the Tail Group of 13. To mitigate the hERG inhibition observed in 13, we explored a number of amide analogs as shown in Table 3. Incorporation of substituents at the ortho-position of the amide carbonyl, such as chloro (16), increased P-gp recognition, while reducing hERG inhibitory activity. The fluoropyridine 17 reduced hERG potential and improved P-gp reflux, although the modification was associated with decreased potency. Replacement of the pyridine with a pyrazine generated compound 18 with a mitigated P-gp efflux and hERG potential. However, these improvements came at the price of decreased microsomal stability in human and increased CYP2D6 inhibition. In an effort to reduce such liabilities, we focused on exploring an alternative to the methoxy group in 18. While the difluoromethyl 19 increased metabolic stability in human microsomes, it lacked potency and still inhibited CYP2D6 with an IC50 value of less than 1 µM. Incorporation of a fluorine on the methoxy led to 20 which had an


11

Page 13 of 82 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


improved potency relative to 19, while retaining favorable values of P-gp efflux and hERG inhibition but also strong CYP2D6 inhibition. Because 20 achieved a reduced P-gp efflux compared to 13, it was evaluated in a PK study with rat. As shown in Table 2, 20 demonstrated an improved B/P ratio of 1.6 as well as retaining favorable PK profiles such as a low CL and moderate to high Cmax and F values. Encouraged by the profile for 20, we considered additional analogs to address the CYP2D6 inhibition.


12


Page 14 of 82

Table 3. Optimization of the Tail Group F H2 N H N

R

O N

O

F F

IC50 (nM)a BACE1b Cellular A c

R

compd

NC

N

13

NC

P-gp ERd

hERG 3 µM (%)e

CYP2D6 IC50 (µM)g

55

110 97

10

43

77 98

>20

1.9

39

95 95

12

1.6

1.9

22

9.4 107

Page 1 of 82 ACS Paragon Plus Environment ... - ACS Publications

Recommend Documents