Structure-Based Drug Design of RN486, a Potent and Selective

Brief Article pubs.acs.org/jmc

Structure-Based Drug Design of RN486, a Potent and Selective Bruton’s Tyrosine Kinase (BTK) Inhibitor, for the Treatment of Rheumatoid Arthritis Yan Lou,* Xiaochun Han, Andreas Kuglstatter, Rama K. Kondru, Zachary K. Sweeney, Michael Soth, Joel McIntosh, Renee Litman, Judy Suh, Buelent Kocer, Dana Davis, Jaehyeon Park, Sandra Frauchiger, Nolan Dewdney, Hasim Zecic, Joshua P. Taygerly, Keshab Sarma, Junbae Hong, Ronald J. Hill, Tobias Gabriel, David M. Goldstein, and Timothy D. Owens pRED, Pharma Research & Early Development, Small Molecule Research, Discovery Chemistry, Hoffmann-La Roche Inc., 3431 Hillview Avenue, Palo Alto, California 94304, United States S Supporting Information *

ABSTRACT: Structure-based drug design was used to guide the optimization of a series of selective BTK inhibitors as potential treatments for Rheumatoid arthritis. Highlights include the introduction of a benzyl alcohol group and a fluorine substitution, each of which resulted in over 10-fold increase in activity. Concurrent optimization of drug-like properties led to compound 1 (RN486) (J. Pharmacol. Exp. Ther. 2012, 341, 90), which was selected for advanced preclinical characterization based on its favorable properties.

■

INTRODUCTION Rheumatoid arthritis (RA) is a serious autoimmune disease with an unknown etiology. Despite the discovery of many efficacious biological agents for the treatment of this condition, there is still a critical need for novel agents with different mechanisms of action for patients unresponsive to current therapies. Furthermore, orally bioavailable small molecules would be particularly desirable in the treatment of RA given both their convenience of administration and ability to clear more rapidly from the body than their biological counterparts in the event of a serious infection during treatment. The most notable recent advance in developing an oral drug for RA was the introduction of tofacitinib,1−4 Pfizer’s selective pan-JAK inhibitor. The successful development of tofacitinib demonstrated that small molecule oral therapies for RA could achieve similar efficacy to biological agents. Additional efforts to introduce novel RA treatments have focused on the development of inhibitors of B-cell function. For example, the SYK inhibitor fostamatinib2,5 from Rigel/ AstraZeneca advanced to phase IIb trials. BTK, a downstream kinase of SYK, has also been shown to play crucial roles in B cell and mast cell activation. The selective BTK inhibitors, PCI32765 (2)6 and CGI1746 (3),7 have shown excellent activity in preclinical animal models for RA. However, proof-of© XXXX American Chemical Society

concept studies in RA patients are yet to be reported. This manuscript describes our structure-based design of compound 1, a potent and selective BTK inhibitor (Chart 1).8

■

CHEMISTRY The synthesis of compound 1 is described in Scheme 1. Intermediate 7 was assembled from nitropyridine 4 through aromatic substitution with N-methyl piperazine, nitro reduction, selective mono Buchwald/Hartwig amination with Nmethyl dibromopyridone, and subsequent conversion to the Chart 1

Special Issue: New Frontiers in Kinases Received: March 3, 2014

A

dx.doi.org/10.1021/jm500305p | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Brief Article

Scheme 1a

essentially the same as that later reported for 3 (Figure 1A).7 Both inhibitors form bidentate hydrogen bonding interactions

a

Reactions and conditions: (a) N-methyl piperizine, K2CO3, Bu4NI, DMSO, 120 °C, 83%; (b) Pd/C, MeOH, rt, 98%; (c) Pd2dba3, Xantphos, Cs2CO3, dioxane, 120 °C, 28%; (d) Pd(OAc)2, X-Phos, pinacolato diboron, KOAc, 100 °C, 84%; (e) carbonyldiimidazole, NH4OH, THF, rt, 87%; (f) Pd2dba3, cyclopropylboronic acid, K2CO3, toluene/water, reflux, 82%; (g) (i) (CH3)2NCH(OCH3)2, MeTHF, 60 °C, (ii) tBuOK, THF, 60 °C, 77%; (h) CuI, K2CO3, DMF, 120 °C, 60%; (i) NaBH4, CH2Cl2/i-PrOH, 4 °C, 87%; (j) 7, Pd(OAc)2, PCy3, K2CO3, dioxane, 88 °C, 83%.

desired boronic ester. The synthesis of intermediate 13 was achieved via the following sequence from benzoic acid 8: (i) amide coupling with ammonium hydroxide, (ii) Suzuki coupling to give cyclopropane 10, (iii) one-pot procedure of converting primary amide in 10 to bicycle 11 through amidination and base-induced ring closure, (iv) selective Suzuki coupling with 2-bromo-6-chlorobenzaldehyde to yield 12, and (v) reduction of the aldehyde group. Finally, compound 1 was assembled through a Suzuki coupling between intermediates 7 and 13.

■

RESULTS AND DISCUSSION Inspired by CGI Pharmaceuticals’ pioneering work on a class of imidazopyrazines specifically claimed as BTK inhibitors7 (i.e., 3, 14), we initiated a crystallography effort aimed at understanding the binding mode of these compounds. No satisfactory docking pose with BTK was obtained using structures available at the time. This work led to the first novel in-house cocrystal structure of 14 (Chart 2) with BTK. The binding mode observed for 14 to BTK kinase domain is

Figure 1. X-ray crystal structures of BTK−inhibitor complexes in balland-stick representation: (A) compound 3 at 1.8 Å resolution (PDB accession number 3OCS); (B) compound 14 at 1.55 Å resolution (4OT5). Inhibitors are displayed in green, selected amino acids in yellow, and water molecules in cyan. Black dashes indicate hydrogen bonds.

with the backbone of Met477 in the kinase hinge region. The tertiary benzamide groups stack between residues Gly480 and Leu408 and point toward solvent. The ligands’ phenyl linker moieties are interacting with the Val416 side chain (not shown) and stack “edge-to-face” against the three consecutive peptide bonds L408-G409-T410-G411 of the Gly-rich loop (amino acids 408−417). The t-butyl-benzamide group of 3 and 14 points into a selectivity pocket formed by the side chains of Q412, F413 (both Gly-rich loop), D521, N526 (both catalytic loop), D539 (DFG), L542, S543, V546, and Y551 (all activation loop). The ligands secondary amide moiety forms a direct hydrogen bond with the side chain of the conserved K430 as well as water-mediated hydrogen bonds with the Glyrich loop.

Chart 2

B

dx.doi.org/10.1021/jm500305p | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Brief Article

high clearance.10 Simply shifting a nitrogen atom from the central bicycle to the aniline ring resulted in compound 18 (Chart 3). The clogP of 18 was half a unit lower (clogP = 5.1)

Interestingly, when the lipophilic t-butyl group is replaced with a much more polar methyl sulfone moiety, the corresponding analogue 15 (Chart 2) no longer occupies the specificity pocket and instead adopts a “U” shape conformation, putting the sulfone in the solvent exposed region (Figure 2B),

Chart 3

than the clogP compound 16. As coplanarity of the aniline ring and the bicyclic group was conserved, 18 has a similar potency (BTK IC50 of 30 nM) to 16. Subsequent truncation of the bicyclic core to a monocyclic pyridone core led to compound 19. With a now substantially reduced clogP of 3.8, not only did we observe an increase in BTK IC50 (10 nM) but 19 also exhibited an improved IC50 of 700 nM in the HWB assay. To confirm that coplanarity of the core and front aniline group is preferred, compound 20 was prepared. This inhibitor was more than 10-fold less potent inhibitor of BTK (IC50 of 160 nM) relative to analogue 19. Despite much reduced lipophilicity, compound 19 was found to be rapidly cleared both in vitro and in vivo. Metabolite identification of 19 in liver microsomes showed that compounds 21 and 22 (Chart 4), arising from oxidation of Chart 4

Figure 2. X-ray crystal structures of BTK−inhibitor complexes in ribbon representation: (A) compound 14 at 1.55 Å resolution (PDB accession number 4OT5); (B) compound 15 at 2.05 Å resolution (4OT6). The sulfone-benzamide moiety of 15 is poorly resolved. Inhibitors (green) and selected amino acid side chains are shown in ball-and-stick representation. Black dashes indicate hydrogen bonds.

the morpholine ring and t-butyl group, are the major metabolites. Many other analogues made without the morpholine group in 19 still suffered from high in vitro and in vivo clearance. Liver microsome metabolite identification experiments consistently showed t-butyl oxidation, sometimes exclusively (i.e., 23). Therefore, many analogues with different replacements of the t-butyl group were prepared and profiled. One of the most interesting analogues prepared in this effort was compound 24, which features a N,N-dimethyl group. Although it is less potent than 20 (BTK IC50 of 20 nM, HWB IC50 4600 nM), it has good pharmacokinetic properties in rats with 36% bioavailability at 5 mg/kg, AUC of 3 μM·h, and a clearance rate of 18 mL/min/kg. With the goal of improving the potency of 20, we analyzed the crystal structure of BTK in complex with 24 for possible ligand modifications that would lead to additional interactions with the protein. The methyl group in the central phenyl group is in close proximity to both K430 and D539 (3.9 Å for C−N of K430 and 3.6 A for C−O of D539). It was envisioned that a hydroxyl substitution on the methyl group might be able to form a hydrogen bond with both K430 and D539. The designed benzyl alcohol analogue 25 (Chart 5) was found to have a Kd of 0.4 nM for BTK as compared with 8 nM for 24. Its

This substitution results in loss of affinity for BTK (IC50 of 1200 nM) and specificity against other kinases (>90% inhibition of 36/256 kinases in Ambit kinome panel at 10 μM concentration).9 A cocrystal structure of an ortho-methyl analogue 16 with BTK was also obtained. Inhibitor 16 (IC50 of 10 nM) is 5-fold more potent than compound 17 (IC50 of 50 nM), which lacks the methyl group on the central phenyl ring. The increased affinity from the methyl group may come from favorable hydrophobic interactions with the protein as well as increased population of the bioactive conformer. Although 16 is a reasonably potent BTK inhibitor in vitro, the combination of suboptimal efficacy in human whole blood (HWB) assays (HWB IC50 for CD69 inhibition >1 μM) and high molecular weight and high lipophilicity (MW 588/clogP 5.6) makes it a fairly unattractive starting point for lead optimization. However, attracted by the exquisite BTK selectivity of this inhibitor and the availability of cocrystal structure information, we decided to engage in issue-driven optimization of the series, focusing on addressing key issues via structure-based optimization. The first issue tackled was the high lipophilicity, which is often associated with low solubility, high protein binding, and C

dx.doi.org/10.1021/jm500305p | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Brief Article

While 24 showed highly desirable PK properties, further identification of metabolites revealed formation of primary aniline metabolite 26. Because of the known mutagenicity risk of many primary anilines,12 compound 26 was tested in Ames and MNT assays and found to be positive in the presence of S9. Because this amide bond is almost in plane with the phenyl group off the carbonyl group and the NH is pointed into solvent, we decided to investigate some bicyclic systems to avoid potential aniline release. First bicyclic quinazolinone 27 (BTK IC50 of 116 nM, Chart 6) was found to be substantially

Chart 5

Chart 6 HWB IC50 improved correspondingly to 120 nM. The increased affinity is attributed to the alcohol’s simultaneous hydrogen bonding interactions with K430 and D539, sometimes referred as cooperative hydrogen bonds. This was confirmed by molecular modeling and subsequent X-ray crystallography (see Figure 3 for a methyl vs hydroxyl−methyl comparison). These interactions are known to significantly boost ligand affinity. 11 Many other substituents were subsequently examined, and all of them were inferior to the benzyl alcohol group in terms of binding affinity.

less active than the corresponding amide 28 (BTK IC50 of 28 nM). After a cocrystal structure of a close analogue 29 with BTK was obtained, it was recognized that the repulsive interaction between the aza group of the new bicycle with N526 may be responsible for the reduction in inhibitor potency (Figure 3A). The reduced form of 27, dihydroquinazolinone 30, was found to be much more potent (BTK IC50 of 50 μM). Subsequent optimization of the bicyclic moiety resulted in cyclopropyl isoquinolone derivative 34, which is devoid of N,Ndimethyl substitution, thus circumventing the potential for formation of primary aniline metabolites upon double demethylation by CYP mediated oxidation. Compound 34 has BTK Kd of 4 nM and HWB IC50 of 104 nM. Fluorine substitution on the benzene ring ortho to the carbonyl group led to 1, which exhibited a significantly improved binding affinity to BTK (Kd of 0.3 nM) in addition to superior HWB potency (IC50 of 17 nM). To better understand the fluorine effect, a crystal structure of a close analogue 35 with BTK was obtained (Figure 3B). The electronegative fluorine atom is within van-der-Waals distance to the primary amine of K430 and the aromatic hydrogen at the ortho position of F413. This suggests that the approximate 10-fold increase in potency gained with the fluorine substituent is likely due to electrostatic interactions with the protein. More thorough SAR studies of the effect of fluorine substitution on different bicyclic moieties will be disclosed in subsequent manuscripts.13 Compound 1 maintains excellent kinase selectivity for BTK after the iterative optimization.1 Despite a less desirable aqueous solubility profile (0.1/1/1222 μg/mL in water, SIF, and SGF, respectively) than 33, compound 1 has a favorable in vivo PK profile with F% of 26/38, AUC of 0.57/2.37 μM·h, and T1/2 of 9.8/11 h in rats and dogs (PO administration, 2 mg/kg). Detailed preclinical biological profiling of 1 has been described elsewhere and demonstrates that this inhibitor has excellent efficacy in various animal models of RA.1 Given its desirable PK and PD profile, compound 1 was advanced into preclinical development studies.

■

■

(1) Xu, D.; Kim, Y.; Postelnek, J.; Vu, M. D.; Hu, D. Q.; Liao, C.; Bradshaw, M.; Hsu, J.; Zhang, J.; Pashine, A.; Srinivasan, D.; Woods, J.; Levin, A.; O’Mahony, A.; Owens, T. D.; Lou, Y.; Hill, R. J.; Narula, S.; DeMartino, J.; Fine, J. S. RN486, a selective Bruton’s tyrosine kinase inhibitor, abrogates immune hypersensitivity responses and arthritis in rodents. J. Pharmacol. Exp. Ther. 2012, 341, 90−103. (2) Fleischmann, R. Novel small-molecular therapeutics for rheumatoid arthritis. Curr. Opin. Rheumatol. 2012, 24, 335−341. (3) Garber, K. Pfizer’s JAK inhibitor sails through phase 3 in rheumatoid arthritis. Nature Biotechnol. 2011, 29, 467−468. (4) Flanagan, M. E.; Blumenkopf, T. A.; Brissette, W. H.; Brown, M. F.; Casavant, J. M.; Shang-Poa, C.; Doty, J. L.; Elliott, E. A.; Fisher, M. B.; Hines, M.; Kent, C.; Kudlacz, E. M.; Lillie, B. M.; Magnuson, K. S.; McCurdy, S. P.; Munchhof, M. J.; Perry, B. D.; Sawyer, P. S.; Strelevitz, T. J.; Subramanyam, C.; Sun, J.; Whipple, D. A.; Changelian, P. S. Discovery of CP-690,550: a potent and selective Janus kinase (JAK) inhibitor for the treatment of autoimmune diseases and organ transplant rejection. J. Med. Chem. 2010, 53, 8468−8484. (5) Singh, R.; Masuda, E. S.; Payan, D. G. Discovery and Development of Spleen Tyrosine Kinase (SYK) Inhibitors. J. Med. Chem. 2012, 55, 3614−3643. (6) Honigberg, L. A.; Smith, A. M.; Sirisawad, M.; Verner, E.; Loury, D.; Chang, B.; Li, S.; Pan, Z.; Thamm, D. H.; Miller, R. A.; Buggy, J. J. The Bruton tyrosine kinase inhibitor PCI-32765 blocks B-cell activation and is efficacious in models of autoimmune disease and B-cell malignancy. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 13075− 13080. (7) Di Paolo, J. A.; Huang, T.; Balazs, M.; Barbosa, J.; Barck, K. H.; Bravo, B. J.; Carano, R. A.; Darrow, J.; Davies, D. R.; DeForge, L. E.; Diehl, L.; Ferrando, R.; Gallion, S. L.; Giannetti, A. M.; Gribling, P.; Hurez, V.; Hymowitz, S. G.; Jones, R.; Kropf, J. E.; Lee, W. P.; Maciejewski, P. M.; Mitchell, S. A.; Rong, H.; Staker, B. L.; Whitney, J. A.; Yeh, S.; Young, W. B.; Yu, C.; Zhang, J.; Reif, K.; Currie, K. S. Specific Btk inhibition suppresses B cell- and myeloid cell-mediated arthritis. Nature Chem. Biol. 2011, 7, 41−50. (8) Lou, Y.; Owens, T. D.; Kuglstatter, A.; Kondru, R. K.; Goldstein, D. M. Bruton’s tyrosine kinase inhibitors: approaches to potent and selective inhibition, preclinical and clinical evaluation for inflammatory diseases and B cell malignancies. J. Med. Chem. 2012, 55, 4539−4550. (9) Karaman, M. W.; Herrgard, S.; Treiber, D. K.; Gallant, P.; Atteridge, C. E.; Campbell, B. T.; Chan, K. W.; Ciceri, P.; Davis, M. I.; Edeen, P. T.; Faraoni, R.; Floyd, M.; Hunt, J. P.; Lockhart, D. J.; Milanov, Z. V.; Morrison, M. J.; Pallares, G.; Patel, H. K.; Pritchard, S.; Wodicka, L. M.; Zarrinkar, P. P. A quantitative analysis of kinase inhibitor selectivity. Nature Biotechnol. 2008, 26, 127−132. (10) Bickerton, G. R.; Paolini, G. V.; Besnard, J.; Muresan, S.; Hopkins, A. L. Quantifying the chemical beauty of drugs. Nature Chem. 2012, 4, 90−98. (11) Kuhn, B.; Fuchs, J. E.; Reutlinger, M.; Stahl, M.; Taylor, N. R. Rationalizing tight ligand binding through cooperative interaction networks. J. Chem. Inf. Model. 2011, 51, 3180−3198. (12) Bentzien, J.; Hickey, E. R.; Kemper, R. A.; Brewer, M. L.; Dyekjaer, J. D.; East, S. P.; Whittaker, M. An in silico method for predicting Ames activities of primary aromatic amines by calculating the stabilities of nitrenium ions. J. Chem. Inf. Model. 2010, 50, 274− 297. (13) Bohm, H. J.; Banner, D.; Bendels, S.; Kansy, M.; Kuhn, B.; Muller, K.; Obst-Sander, U.; Stahl, M. Fluorine in medicinal chemistry. ChemBioChem 2004, 5, 637−643.

ASSOCIATED CONTENT

* Supporting Information S

Experimental procedures; 1HNMR, HRMS spectra. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: 408-332-6502. E-mail: [email protected] or ylou@ nurix-inc.com. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge Xingrong Liu, Jennifer Fretland, Bill Fitch, and Bo Wen for providing drug metabolism support and Naina Patel and Qingyan Hu for formulation support. This work was supported exclusively by Hoffmann-La Roche.

■

ABBREVIATIONS USED BTK, Bruton’s tyrosine kinase; CYP, cytochrome P450 enzymes; hERG, human ether-a-go-go related gene; HWB, human whole blood; JAK, Janus kinase; MNT, micronucleus test; RA, rheumatoid arthritis; S9, supernatant fraction obtained from liver homogenate by centrifuging at 9000g for 20 min in a suitable medium, this fraction contains cytosol and microE

dx.doi.org/10.1021/jm500305p | J. Med. Chem. XXXX, XXX, XXX−XXX