A scaffold-hopping approach to discover potent, selective and

Subscriber access provided by - Access paid by the | UCSB Libraries

Article

A scaffold-hopping approach to discover potent, selective and efficacious inhibitors of NF-#B inducing kinase Nicole Blaquiere, Georgette Castanedo, Jason Burch, Leo Berezhkovskiy, Hans Brightbill, Suzanne Brown, Connie Chan, Po-Chang Chiang, James J. Crawford, Teresa Dong, Peter Fan, Jianwen Feng, Nico Ghilardi, Robert Godemann, Emily Gogol, Alice Grabbe, Alison Hole, Baihua Hu, Sarah G Hymowitz, Moulay Hicham Alaoui Ismaili, Hoa Le, Patrick Lee, Wyne Lee, Xingyu Lin, Ning Liu, Paul McEwan, Brent McKenzie, Hernani Silvestre, Eric Suto, Swathi Sujathabhaskar, Guosheng Wu, Lawren C. Wu, Yamin Zhang, and Steven T Staben J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00678 • Publication Date (Web): 25 Jun 2018 Downloaded from http://pubs.acs.org on June 25, 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 45

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

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 45

A scaffold-hopping approach to discover potent, selective and efficacious inhibitors of NF-κB inducing kinase Nicole Blaquiere*,†, Georgette M. Castanedo†, Jason D. Burch†, Leonid M. Berezhkovskiy†, Hans Brightbill†, Suzanne Brown†, Connie Chan†, Po-Chang Chiang†, James J. Crawford†, Teresa Dong†, Peter Fan†, Jianwen Feng†,¦, Nico Ghilardi†, Robert Godemann‡, Emily Gogol†, Alice Grabbe‡, Alison J. Hole‡, Baihua Hu+, Sarah G. Hymowitz†, Moulay Hicham Alaoui Ismaili†, Hoa Le†,¤, Patrick Lee†, Wyne Lee†, Xingyu Lin+, Ning Liu†, Paul A. McEwan‡, Brent McKenzie†, Hernani L. Silvestre‡, Eric Suto†, Swathi Sujatha-Bhaskar†, Guosheng Wu+, Lawren C. Wu†, Yamin Zhang+ and Steven T. Staben*,† †

+

‡

Genentech, Inc. 1 DNA Way, South San Francisco, CA 94080

Pharmaron Beijing Co., Ltd. 6 Taihe Road, BDA, Beijing, P.R. China, 100176

Evotec AG, Manfred Eigen Campus, Essener Bogen, Hamburg, Germany 22419

ABSTRACT

NF-κB-inducing kinase (NIK) is a protein kinase central to the non-canonical NF-κB pathway downstream from multiple TNF receptor family members, including BAFF, which have been


1

Page 3 of 45 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


associated with B cell survival and maturation, dendritic cell activation, secondary lymphoid organ development, and bone metabolism. We report herein the discovery of a lead chemical series of NIK-inhibitors that was identified through a scaffold-hopping strategy using structurebased design. Electronic and steric properties of lead compounds were modified to address glutathione conjugation and amide hydrolysis.

These highly potent compounds exhibited

selective inhibition of LTβR-dependent p52 translocation and transcription of NF-κB2 related genes. Compound 4f is shown to have a favorable pharmacokinetic profile across species and to inhibit BAFF-induced B cell survival in vitro and reduce splenic marginal zone B cells in vivo.

INTRODUCTION NF-κB inducing kinase (NIK) is a key regulator of non-canonical NF-κB signaling.1 NIK is a ‘non-RD’ (non-arginine-aspartate) kinase2 and activation of NIK downstream of TNF receptor superfamily members (such as BAFF, CD40 and TWEAK) is achieved by regulation of NIK’s stability. In the absence of pathway stimulation, co-binding of NIK, TRAF2/3 and cIAP lead to cIAP-mediated ubiquitination and proteasomal degradation of NIK.3 Upon stimulation, TNF receptor superfamily members are thought to bind the TRAF2/3 – cIAP complex, decreasing the proximity of NIK to cIAP and thus improving the stability of NIK.

Accumulated and

constitutively active NIK phosphorylates and activates IKKα, and downstream events lead to p100 processing, nuclear translocation of p52, and increased transcription of NF-κB target genes.4 Several TNF family superfamily members have been implicated in systemic lupus erythematosus (SLE) pathophysiology, most notably BAFF,5 CD40 and TWEAK.6 We hypothesize that inhibiting the kinase function of NIK would target common downstream pathways of these implicated tumor necrosis factor receptor superfamily (TNFRSF) receptors


2


Page 4 of 45

and potentially provide superior efficacy in SLE compared to targeting individual TNFRSF receptors with biologic approaches.7 Several efforts have been disclosed detailing small molecule inhibitors of NIK,8 but few have proved to generate appropriately potent and selective chemical tools to prosecute the relevance of NIK activity in auto-inflammatory diseases. We recently reported a structure-based design effort leading to a series type 1.5 inhibitors of NIK possessing a benzoxepin core structure and pendant propargyl alcohol (Figure 1, compounds 1).9 While large improvements in potency and kinase selectivity were obtained, these compounds were insufficient for robust in vivo evaluation of NIK pharmacology. Here we describe a scaffold-hopping approach10 directed at further improvements in cell potency, kinase selectivity and in vivo PK properties. RESULTS AND DISCUSSION A minimalist approach was taken to chemically simplify and identify new vectors of substitution within the benzoxepin series. Our previous studies indicated the benzoxepin oxygen atom was not required for NIK potency and could be replaced with a bridged bicyclic carbon core, yielding improved selectivity over several kinases including phosphatidylinositol-3-kinase (PI3K).9 We also reported some substituents on the 5-membered aromatic heterocycle afforded improved kinase selectivity without much consequence to NIK affinity. With this in mind, we hypothesized that removal of the oxepin ring would be tolerated, simplifying the synthetic accessibility of compounds with substitution on the heterocyclic core (Figure 1). A crystal structure of representative benzoxazepin 1a in murine NIK11 is shown in Figure 1 (PDB ID 6G4Y). The core positions a primary amide to accept and donate hydrogen bonds to hinge residues Leu474 and Glu472, as well as directs the propargyl alcohol substituent past the gatekeeper Met471 to engage critical residues common for type 1.5 inhibitors (αC-helix Glu442


3

Page 5 of 45 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


and DFG-motif Phe537).12 In order to maintain these key interactions, as well as reinforce optimal binding conformation created by the constrained rigidity of the oxepin, we first designed 1-(2-pyridyl)-3-amido pyrazoles exemplified by compounds 2. We anticipated a desirable torsional preference of the primary amide and biaryl C-N bond based on an intramolecular hydrogen bond and electron pair repulsion, respectively. This design was also anticipated to allow rapid access to multiple analogs using simple SNAr chemistry (vide infra). We chose 3ethynyl-3-hydroxy-1-methylpyrrolidin-2-one substitution at the 4-position of the pyridine given reliably high lipophilic ligand efficiency (LLE)13 analogs created in the benzoxepin series.9 (a) O

Me

NH 2

Me

NH 2

O

N

N

O

N X X X

OH

N N

R1 R2

O 1

O OH N 2

(b)


4


Page 6 of 45

Figure 1. (a) Proposed excision of oxepin ring of 1 to 2-(3-amido)pyrazolyl-pyridines 2. Crystal structure of benzoxazepin 1a (blue) in murine NIK (gray) (PDB ID 6G4Y). Solvent accessible surface indicated in gray. Key intermolecular distances are indicated with gray lines/text. Biochemical activity and LLE values for initial ring-opened derivatives are displayed in Table 1. The ability of compounds to inhibit recombinant NIK protein was determined by measuring inhibition of the native rate of substrate-independent hydrolysis of ATP to ADP (“NIK ADP FP”).14 Cellular activity was determined through inhibition of the expression of the NF-κB response element-regulated firefly luciferase reporter in HEK-293 cells (dox-inducible NIK, “NF-κB reporter”).9 As we have demonstrated previously, large shifts between inhibition in biochemical and cell based assays emphasize the need for exquisitely potent inhibitors. We believe this to be partially due to NIK’s low Km for ATP (~7 µM).9 Unsubstituted pyrazole 2a is a moderately potent inhibitor in the biochemical assay (Ki of 28 nM) and expectedly does not influence the NF-κB reporter assay at concentrations up to 20 µM. This compound, however, does have a desirable LLE (6.9) based on a low measured LogDpH 7.4 (0.61). We hypothesized increasing the LogD of the compound with non-polar substitution off the pyrazole would increase the inhibitory potency by aiding desolvation of the compound. The addition of small lipophilic substituents to the 5-position of the pyrazole ring such as compound 2c did indeed improve cell potency to the single-digit µM range, as did the addition of an unsubstituted amino group (2b). We hypothesized the amine in 2b could form an intramolecular hydrogen bond with the pyridine nitrogen, providing stabilization of the low energy binding conformation. Cellular activity of this analog, however, was modest. Fused bicyclic systems exemplified by compounds 2d-f proved stronger inhibitors of NIK activity. In particular, indazoles such as 2e and 2f afforded impressive potency gains. Indazole 2e (Ki = 0.08 ± 0.05 nM; cellular IC50 = 33 ± 9


5

Page 7 of 45 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


nM) is >250-fold more active than pyrazole 2a, maintains a desirable overall lipophilicity (LogDpH 7.4 = 1.9) and improves LLE substantially (8.2 vs. 6.9, respectively). Table 1: Biochemical and cellular inhibition data for initial ring-opened analogs.

R1

Compound

NIK ADP-FP Ki (nM)a

NF-κB reporter IC50 (µM)b

LogDpH 7.4 c / LLEd

2a, R2 = H

28 ± 18 (n=2)

>20 (n=2)

0.61 / 6.9

2b, R2 = NH2

9.4 ± 5 (n=2)

4.3

2.0 / 6.0

2c, R2 = cPr

16

2.1

2.0 / 5.8

2d

0.86 ± 0.34 (n=4)

0.084 ± 0.037 (n=2)

1.9 / 7.2

2e, R3 = H

0.08 ± 0.05 (n=4)

0.033 ± 0.09 (n=3)

1.9 / 8.2

2f, R3 = F

0.17 ± 0.13 (n=2)

0.22 ± 0.04 (n=2)

N.R.

N.R. – not reported. Data are represented as mean ± standard deviation of the average for the number (n) of experiments conducted. aInhibition of NIK-catalyzed hydrolysis of ATP to ADP (FP = fluorescence polarization). bInhibition of expression of NF-κB response element-regulated firefly luciferase reporter gene in HEK293 cells. cLogD measured using a high-throughput microscale shake flask and liquid chromatography tandem mass spectrometry approach.9 d Lipophilic ligand efficiency (LLE) was calculated utilizing measured LogD values via the following equation: [NIK ADP-FP pKi] – [LogDpH 7.4]. Compounds 2d-f were tested for single point inhibition of select additional kinases at a concentration of 100 nM, >100x the biochemical Ki for NIK. These compounds were determined


6


Page 8 of 45

to have high selectivity for NIK; only 1 or 2 kinases (KHS1 and PKD1) were inhibited at > 50 % (Figure 2a, tabulated data in supplemental information). A co-crystal structure of 2f bound to murine NIK confirmed our proposed binding mode of this series (Figure 2b, PDB ID 6G4Z). As designed, we observed that 2f (blue sticks) aligns well with benzoxazepin 1a (magenta wireframe); the primary amide lies in proximity of the hinge to interact with residues Leu474 and Glu472, and the propargyl alcohol maintains critical H-bonding distances with the DFGmotif and αC-helix glutamate. From this structure it is also clear the ligand occupies an area of the active site (between the lower hinge and P-loop) that was otherwise difficult to access with the benzoxepin series. Interestingly, the observed dihedral angle between aryl rings of the small molecule when bound to murine NIK (mNIK) is ~19º out of plane, in opposition to the ~0º dihedral angle observed with benzoxazepins such as 1a. We hypothesized that the increased rotational freedom around this bond may allow for more optimal space filling and hydrogen bonding interactions of the primary amide and propargyl alcohol, accounting for some of the observed inhibitory potency. (a)

(b)


7

Page 9 of 45 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. (a) Inhibition of kinase activity at a single concentration (100 nM) across a range of kinases tested at Invitrogen® (red >80%-inhibition, yellow 50 - 80%, green < 50%). (b) X-ray co-crystal structure of 2f (blue) in mNIK (gray) including an overlay of benzoxazepin 1a (magenta wire-frame) (PDB ID 6G4Z). Solvent accessible surface indicated in gray. Key intermolecular distances are indicated with gray lines/text. (a)

(b) Compound 2d 2e 2f

HH CLhep HLM CLhep (mL/min/kg) (mL/min/kg) 1.6 14.9 7.7 15.2 8.6 14.1

Figure 3. (a) Human liver microsome (HLM) and human hepatocyte (HH) predicted hepatic clearance.14 (b) Qualitative representation of metabolism pathways observed in HH after 3 hours incubation with 5 µM 2f.14 Compounds 2d-f were incubated with human liver microsomes (HLM) and human hepatocytes (HH) to evaluate their metabolic stability (Figure 3a).14 Compound 2d was particularly stable in HLM hepatic clearance of 1.6 mL/min/kg. The other two molecules were found to be moderately stable in HLM, with predicted hepatic clearance between 30% and 50% liver blood flow. In contrast, all three compounds were rapidly metabolized in HH (predicted hepatic clearance > 14 mL/min/kg, > 70% liver blood flow) suggesting additional phase II metabolism pathways may be operative. Metabolites were analyzed after 3 hour incubation of 2f with human hepatocytes. This experiment revealed that a major metabolite resulted from glutathione (GSH) addition to the parent molecule. We hypothesized 1,6 addition of glutathione to the electron-deficient alkyne may be possible given the electron-withdrawing nature of the 4-pyridyl (schematic in Figure


8


Page 10 of 45

2b).15 This glutathione conjugation was not observed in the benzoxepin series of NIK inhibitors. The other major route of metabolism was hydrolysis of the primary amide, presumably catalyzed by the action of hydrolases present in the hepatocyte prep.16

Table 2. Biochemical and cellular potency and metabolic stability data for select analogs

R2

NIK ADPFP Ki (nM)

NF-κB reporter IC50 (nM)

HLM / HH CLhep (mL/min/kg)

LogDpH 7.4 / LLE

3a

-

350

n/a

4.8 / -

2.4 / 4.1

3b

H

0.23 ± 0.23 (n=2)

38 ± 11 (n=21)

7.4 / 10

1.6 / 8.0

3c

OMe

0.13 ± 0.14 (n=3)

21 ± 5 (n=4)

11 / 11

1.9 / 8.0

3d

H

0.23 ± 0.23 (n=2)

29 ± 9 (n=4)

7.5 / 13

2.1 / 7.5

3e

F

0.27± 0.05 (n=4)

39

11 / 7.3

2.6 / 7.2

3f

H

4.7 ± 0.21 (n=2)

500 ± 64 (n=2)

4.0 / 2.5

1.3 / 7.0

3g

OEt

5.9 ± 1.4 (n=2)

380 ± 57 (n=2)

16 / -

1.9 / 6.3

Compound

R1


9

Page 11 of 45 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


3h

H

0.18 ± 0.09 (n=2)

15 ± 5 (n=2)

13 / 14

2.2 / 7.7

3i

OMe

0.05 ± 0 (n=2)

4.7 ± 2.7 (n=2)

16 / 12

2.4 / 7.9

See Table 1 for assay details We next synthesized and tested analogs similar to 2e,f where the central pyridine ring was replaced with a 1,3-disubstituted benzene and alternative 6,5 aromatic heterocycles (Compounds 3, Table 2). We anticipated these two changes would minimize the electrophilicity of the alkyne and potentially modulate the rate of primary amide hydrolysis. Importantly, the central pyridyl nitrogen did not appear to be crucial for activity as many analogs containing 6,5 heterocyclic systems were potent in both the biochemical and cell-based assays (Table 2). Among changes investigated, one not tolerated was in azaindole 3a, possibly owing to steric clash of the C-H bond with the methionine gatekeeper and/or to effects on the conformation of the hinge-binding primary amide. Pyrazolopyridines 3b,c were potent and moderately stable in HLM as well as HH, with small substituents such as methoxy having minimal effects on these parameters. Indazole 3d maintained similar potency compared to compound 2e with the pyridyl central ring. This, in combination with an observed biaryl torsion of ~19º for 2f shown previously when bound to mNIK, implies that our initial concerns around controlling the dihedral angle to match the benzoxazepin series were not well-founded. Imidazopyridines were designed as indazole replacements, maintaining the necessary nitrogen placement to control the conformation of the primary amide.

Imidazopyridine 3f was

significantly less potent than indazole comparator 3d (NF-κB IC50 = 500 ± 64 nM vs. 29 ± 9 nM, respectively). However, this compound had promising in vitro stability, with a predicted hepatic clearance of 2.5 mL/min/kg determined in HH.

The alternative imidazopyridine isomer,


10


Page 12 of 45

however, was significantly more potent. With a NIK Ki = 0.05 nM and NF-κB IC50 = 4.7 ± 2.7 nM, 3i was the most potent compound from this set.

Unfortunately, despite favorable

lipophilicity (LogD pH 7.4 < 3), all compounds including this imidazopyridine isomer were labile in HLM and HH. Compound 3d was selected for further profiling due to its favorable potency profile and moderate clearance in HLM and HH. Metabolites were analyzed after 3 hour incubation with HH.14 GSH conjugation to the parent molecule was greatly reduced, which we anticipate to be a result of a less electron withdrawing central phenyl ring. However, the total rate of clearance of 3d (HH CLhep = 13 mL/min/kg) remained a concern. The major metabolite observed from this study was amide hydrolysis. We next put effort toward a further reduction in clearance and increased potency by expansion of the 5-membered heteroaromatic ring. We hypothesized the enzymatic hydrolysis of the primary amide could be altered by modifying the adjacent electronic, steric and/or polar environment.

In addition, ligand-based overlays predicted expansion of the 5-membered

heteroarene would move this key primary amide in more optimal H-bonding distance to the hinge. Data for these analogs (compounds 4) are presented in Table 3. Replacement of the core with a simple phenyl ring was not well-tolerated (4a, NIK Ki = 68 nM, LLE = 5.6) suggesting non-optimal low energy torsions at the biaryl and/or amide C-C bonds.

Inclusion of 2,5-

subsituted pyrazine and 2,4-substituted pyrimidine in analogs 4b and 4c resulted in single-digit nM values in the NIK biochemical assay, however these did not have the desired level of potency in the cell-based NF-κB reporter assay. The isomeric 2,4-substituted pyrimidine and 2amido pyridines 4d-f had improved overall profiles.


11

Page 13 of 45 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


Analogs with these substructures consistently inhibited NIK in the biochemical assay with Ki values < 1 nM. Among these, 4-methoxy-2-amido pyridine 4f had the best cell-based activity (NF-kB reporter IC50 = 34 ± 6 nM). Notably, predicted hepatic clearance of 4f from in vitro HLM and HH experiments was improved relative to above indazole 3d. Additional potency was achievable within this sub-series, albeit at the cost of kinase selectivity. Quinazoline 4g and 3amino-2-amido pyridine 4h were particularly potent in the NF-κB reporter assay (IC50 = 10 ± 5 nM and 16 ± 5 nM, respectively). However, these inhibitors lost significant selectivity against the kinome.14 Compound 4f possessed the combination of cell-based potency and improved liver microsome stability to profile further for progression to in vivo studies. Table 3. Data for compounds 4 with 6-membered R1 substitution.

R

NIK ADP-FP Ki (nM)

NF-κB reporter IC50 (nM)

HLM / HH CLhep (mL/min/kg)

LogDpH 7.4 / LLE

4a

-

68

ND

5 / ND

1.6 / 5.6

4b

-

1.3 ± 0.5 (n= 2)

170 ± 71 (n= 2)

6.1 / 17

1.0 / 7.9

Compound

Core


12


Page 14 of 45

4c

-

2.0 ± 1.1 (n= 2)

355 ± 191 (n= 2)

-

1.6 / 7.1

4d

-

0.59 ± 0.42 (n= 2)

69 ± 11 (n= 2)

6.9 / 8.5

1.8 / 7.4

4e

H

1.6 / 7.8

OMe

80 ± 4 (n= 4) 34 ± 6 (n= 7)

7.1 / 2.9

4f

0.42 ± 0.47 (n= 3) 0.23 ± 0.17 (n= 2)

6.5 / 5.8

1.9 / 7.7

4g

-

0.07 ± 0.03 (n= 2)

10 ± 5 (n= 3)

10 / ND

2.5 / 7.7

4h

-

0.07 ± 0.02 (n= 2)

16 ± 5 (n=3)

10 / 5.6

1.7 / 8.5

See Table 1 for assay details Table 4. In vitro profile of compound 4f

Compound

4f

NIK ADP-FP Ki (nM) / LLE

0.23 ± 0.17 (n= 2) / 7.7

NFκB-LUC IC50 (µM)

0.034 ± 6 (n= 7)

HeLa p52 / HeLa RelA (µM)a

0.070 ± 0.048 (n= 5) / >20 (n= 8)

MW / LogD / TPSA

365 / 1.9 / 105


13

Page 15 of 45 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


a

LM CLhep (H/R/M/D/C)b

6.5 / 21 / 53 / 16 / 19

Hep CLhep (H/R/M/D/C)b

5.8 / 18 / 38 / 21 / 17

Kin Solubility (µM)

139

MDCK (A to B) / (B to A)c

7.30 / 10.25

Plasma protein binding % (H/R/M)

84.9 / 90.1 / 95.3

See supplemental section for experimental details; bUnits of mL/min/kg; cUnits of 10-6cm/s

Additional data for compound 4f are dislpayed in Table 4. Consistent with expectations for a NIK inhibitor, 4f was shown to inhibit nuclear translocation of p52 (RelB) (IC50 = 70 nM). Importantly, nuclear translocation of canonical NF-κB marker RelA was not influenced up to 20 µM test concentration, suggesting high selectivity for inhibition of non-canonical NF-κB signaling.4 Kinase selectivity was very high as 4f inhibited only KHS1 > 50% in a 228 panel of kinases tested at 0.100 µM (>400-fold NIK Ki, see supplemental section). Moderate to good hepatic clearance was predicted across species based on incubation with liver microsomes and hepatocytes. This molecule had a high measured kinetic solubility (139 µM) and moderate passive permeability as determined against a Madin-Darby canine kidney line (MDCK, Papp A-B = 7.3 x 10-6cm/s) which predicted high bioavailability upon oral dosing. In a metabolite ID experiment, amide hydrolysis was substantially reduced for 4f relative to compound 3d and GSH conjugation was not observed.14 Due to these favorable in vitro properties, 4f was further profiled in pharmacokinetic studies. In vivo clearance values generally correlated well with predicted clearance in the microsomal and hepatocyte assays (Table 5). Moderate to excellent bioavailability was observed across species.


14


Table 5. In vivo PK data for 4f across multiple species

SD Rat

CD-1 Mouse

Beagle

Cynomologous Monkey

intravenous Clp (mL/min/kg) Vd (L/kg)

20

32

18

7.8

1.35 susp. in corn oil

0.778 susp. in MCT 56

1.39

formulation

1.58 susp. in MCT 54

a

F (%)a Dosed orally at 5 mg/kg.

97

susp. in MCT >100

We also examined the pharmacology of compound 4f in isolated mouse and human primary Bcells (Figure 4). As expected, compound 4f inhibits BAFF stimulated B cell survival in mouse B-cells in a concentration dependent manner (IC50 = 373 nM). Human B cell proliferation was similarly affected.14 B cell Survival (mouse) 2.5

IC

50

= 373 +_ 64 nM

BAFF (10ng/ml) + Compound 4f BAFF (10 ng/mL) + DMSO

2.0

DMSO only

6

RLU (x10 )

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 16 of 45

1.5 1.0 0.5 0 -12

-10

-8

-6

-4

-2

Compound 4f (µM)

Figure 4. Compound 4f inhibits BAFF-induced B cell (mouse) survival in vitro. Splenic mouse B cells were cultured with mouse rBAFF for four days. B cell survival was measured by Cell Titer Glo (Promega). Mean and standard deviation of 3 biological replicates are shown.


15

Page 17 of 45 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 5. Compound 4f inhibits BAFF signaling in vivo. (A-B) NIK inhibition reduces marginal zone B cell (B220+ CD5- CD21high CD23low) frequency (A) and number (B). (C) Terminal free plasma concentration of Compound 4f. IC50, IC70 and IC90 levels were determined based on in vitro mouse B cell survival assay results (Figure 4). Data represented as mean ± standard deviation of 4-5 mice per group. Statistics: unpaired t-test with Welch’s correction (GraphPad Prism Software; * p

A scaffold-hopping approach to discover potent, selective and

Recommend Documents