1H-Pyrazolo[3,4-b]pyridine Inhibitors of Cyclin-Dependent Kinases: Highly Potent 2,6-Difluorophenacyl Analogues

Raj N. Misra,* Hai-yun Xiao, David B. Rawlins, Weifang Shan, Kristen A. Kellar, Janet G. Mulheron,† John S. Sack, John S. Tokarski,

S. David Kimball† and Kevin R. Webster.

Abstract—Structure–activity studies of 1H-pyrazolo[3,4-b]pyridine 1 have resulted in the discovery of potent CDK1/CDK2 selec- tive inhibitor 21h, BMS-265246 (CDK1/cycB IC50=6 nM, CDK2/cycE IC50=9 nM). The 2,6-difluorophenyl substitution was cri- tical for potent inhibitory activity. A solid state structure of 21j, a close di-fluoro analogue, bound to CDK2 shows the inhibitor resides coincident with the ATP purine binding site and forms important H-bonds with Leu83 on the protein backbone.

Cyclin-dependent kinases (CDKs) are a family of pro- tein kinases which along with their regulatory subunit cyclins play a key role in the growth, development, proliferation and death of eukaryotic cells and are responsible for insuring integrity in the coordination of events through the cell cycle.1 Due to both their central role in the cell cycle and their misregulation in a number of cancers, CDKs have been implicated as contributing factors to the development of cancer. Consequently, oncology drug discovery programs have directed major efforts towards the identification of small molecule inhi- bitors of CDKs as potential therapeutic agents.2 Our screening efforts recently revealed that pyrazolo[3,4- b]pyridines SQ-67563 (1) and SQ-67454 (2) as CDK inhibitors with relatively potent enzyme inhibitoryactivity and selectivity for the CDK family (Fig. 1).3—5Described herein are the synthesis and SAR of the 5-phenacyl substituent and the solid-state structure of an optimized inhibitor of this chemotype bound to CDK2.Acyl analogues (Tables 1 and 2) were prepared via addition to key intermediate amide 9. The overall syn- thetic route is shown in Scheme 1. Thus, neutralization

*Corresponding author. E-mail: [email protected]
†Current address: Lexicon Pharmaceuticals, 350 Carter Road, Prince- ton, NJ 08540, USA.
{Current address: Astra-Zeneca R&D Boston, 35 Gatehouse Dr., Waltham, MA 02451, USA.
of PMB-protected aminopyrazole 35 and condensation with commercial acrylate 4 followed by pyrolysis in diphenyl ether solution (240 ◦C) afforded annulated pyrazole 5. Treatment of 5 with neat phosphorous oxy- chloride gave 4-chloropyrazolopyridine 6. Introduction of the 4-butyloxy substituent via the alkoxide followed by ester hydrolysis afforded acid 7. Acid chloride cou- pling of 7 with N,O-dimethylhydroxylamine followed by removal of the PMB group with TFA yielded key
intermediate 9. In general, treatment of 9 with excess (5–10 equiv) of the appropriate organolithium afforded the desired 5-keto substituted analogues 10 and 21 (Tables 1 and 2).6,7

In contrast to organolithium reagents, addition of Grignard reagents to 9 gave competing substitution for the 4-butoxy group. Attempts to add aryllithiums to PMB-protected intermediate 8 also resulted in pre- dominately substitution for the 4-butyloxy group. Amido analogues 10h and 10i were available from acid 7 as shown in Scheme 2 by generation of the acid chlo- ride, addition of appropriate amine followed by removal of the PMB group with hot TFA.

The 5-thio analogues4d were prepared as shown in Scheme 3 from bicyclic pyridinol ester 5. Hydrolysis of the ester gave acid 12 which upon heating smoothly decarboxylated to afford pyridinone 13. Treatment of 13 with bromine gave aryl bromide 14. The 4-butoxy group group was introduced by chlorination of 14 with

0960-894X/03/$ – see front matter Ⓒ 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0960-894X(03)00381-0

Image2406 R. N. Misra et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2405–2408
Figure 1. Structure of pyrazolo[3,4-b]pyridine screening hits SQ-67563
(1) and SQ-67454 (2).

POCl3 followed by stirring the intermediate 4-chloro analogue 15 with sodium n-butoxide to give 16. The 5-thio substituent was introduced by low temperature metal–halogen exchange of 16 and quenching the resulting anion with diphenyl disulfide to provide 17. TFA removal of the PMB group afforded 5-thiophenyl analogue 18. Oxidation of 18 with 1 equiv of mCPBA provided the racemic 5-sulfoxide 19 while oxidation with excess mCPBA gave and 5-sulfone analogue 20.
Table 1 shows the effect of replacement of the phenyl group in the 5-position on in vitro CDK inhibitory activity in a cell-free enzyme assay.8 Thus, small alkyl groups (10a,b) or benzyl (10c) resulted in >10-fold loss of CDK2 potency. Replacement of the phenyl group with a pyridyl ring (10d), a small basic heterocycle, also resulted in a ~ 200-fold decrease in potency. In contrast, small neutral heterocycles, such as furan, thiophene and thiazole (10e–g), afforded analogues with CDK1 and
CDK2 inhibitory potency comparable to phenyl analo- gue 1. An amide or methyl ester functionality (10h–j) at the 5-position resulted in a >10 to 200-fold loss in CDK2 inhibitory potency. Interestingly, both racemic sulfoxide 19 and sulfone 20 were comparatively potent inhibitors of CDK2 with IC50=0.34 and 0.52 mM, respectively, indicating that a sulfone or sulfoxide may serve as a surrogate for the carbonyl group.

Shown in Table 2 is an examination of phenyl ring sub- stituent effects on CDK inhibitory potency. Thus, sub- stitution of a methyl group in the ortho or meta position of 1 (21b,c) gave a 6- to 7-fold loss in CDK2 inhibitory potency, while para methyl substitution (21d) resulted in only a modest 2-fold loss in CDK2 inhibitory potency relative to 1. In contrast, mono- and di-ortho fluoro substitution (21e and 21f) both resulted in 2- to 3-fold increases in CDK2 inhibitory potency relative to 1. Para

ImageR. N. Misra et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2405–2408 2407

Scheme 3. Synthesis of 5-thio analogues 18–20: (a) aq NaOH/EtOH, 95 ◦C, 94%; (b) 230 ◦C, 15 min, 100%; (c) Br2, EtOH, 0 ◦C, 97%; (d)
POCl3, 110 ◦C, 1 h, 95%; (e) nBuONa/nBuOH, 60 ◦C, 2 h, 86%; (f)
nBuLi/THF, —78 ◦C, 30 min then PhSSPh, 29%; (g) TFA, 65 ◦C, 2.5 h, 62%; (h) mCPBA (1 equiv)/CH2Cl2, 0 ◦C, 74% for 19 or mCPBA(3
equiv)/CH2Cl2, 25 ◦C, 78% for 20.substitution coupled with 2,6-difluoro substitution (21g–k) afforded analogues that were ~ 3- to 12-fold more potent CDK2 inhibitors than 1. In particular, the 2,6-difluoro-4-methylphenyl analogue 21h exhibited CDK1 and CDK2 potency that was 25- and 11-fold more potent versus CDK1 and CDK2, respectively, than phenyl analogue 1 and represented the most potent CDK/CDK2 selective analogue from this chemotype. In contrast to 21h, the 2,4,6-trimethylphenyl analogue 21lwas a very poor inhibitor of CDK2 indicating that the ortho fluoro substituents were a key for potent inhibi- tory activity. In an ovarian cancer cell line (A2780) inhibitor 21h produced a cytotoxic effect with an IC50=0.76 mM.8

Top: Solid-state structure of pyrazolopyridine 21j bound in the ATP-pocket of CDK2 (no cyclin). The inhibitor carbon atoms are colored green, the nitrogen atoms are colored blue and oxygen atoms are colored red. The protein carbons are colored gray. Hydrogen bonds are shown by the magenta dotted lines. Bottom: Overlay of compound 1 (magenta) and 21j (green) bound to CDK2 showing location of 5-phenacyl substituent. The nitrogen atoms are colored blue, the oxygen atoms are colored red, the fluorine atoms are colored slate and the bromine atom is colored in brown.

The three-dimensional structure of 2,6-difluoro analo- gue 21j in complex with CDK2 was determined by X-ray crystallography. Crystals were obtained by incu- bating inhibitor 21j (72 h) with crystalline protein in the absence of cyclin.8 The crystal structure (Fig. 2) revealed that 21j binds in the ATP-binding site as seen previously with 1 forming important hydrogen bonds between the pyrazolopyridine ring and Leu83.

The 4-butoxy substituent extends into the space occu- pied by the ribose of ATP and does not appear to form any specific contacts with the protein while the pendent difluorophenyl ring lies buried within the protein, stacking with Phe80. An overlay of inhibitors 1 and 21j clearly show similar binding modes (see Fig. 2, bottom).

In this arrangement small, flat pendent aryl groups would be expected to bind optimally to the protein while larger groups would create unfavorable steric interactions with Phe80. This is consistent with the SAR shown in Table 1 in which benzyl (10c) and amide sub- stitution (10h and 10i) showed diminished activity. Interestingly, although the pendent phenyl rings of 1 and 21j occupy nearly the same region in space the car- bonyl groups appear to Abe directed orthogonally. The contrasting orientation of the carbonyl group indicates that it is unlikely that the carbonyl oxygen is involved in a H-bonding interaction.

The 2,6-difluorophenyl ringlies nearly coplanar with the carbonyl group (torsional angle=7.4◦) which may provide an explanation for the2408 R. N. Misra et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2405–2408educed activity of the 2,4,6-trimethylphenyl analogue 21l which requires significant skewing of the phenyl and carbonyl groups.In summary, SAR studies of pyrazolopyridine 1 have resulted in the discovery of potent, CDK1/CDK2selective inhibitor 21h (BMS-265246). The 2,6-difluoro- phenyl substitution pattern was critical for potent inhi- bitory activity.

We thank Dr. Bang-Chi Chen and Mr. Mark S. Bed- narz for preparation of chemical intermediates and Dr. John T. Hunt for scientific guidance.
References and Notes

1. (a) Hunter, T.; Pines, J. Cell 1994, 79, 573. BMS-265246 (b) Sherr, C. Science 1996, 274, 1672. (c) Review: Pines, J. Semin. Cancer Biol. 1994, 5, 305.

2. (a) Webster, K. R. Exp. Opin. Invest. Drugs 1998, 7, 1. (b) Webster, K. R.; Kimball, S. D. Emerg. Drugs 2000, 5, 45. (c) Sielecki, T. M.; Boylan, J. F.; Trainor, G. L. J. Med. Chem. 2000, 43, 1. (d) Kimball, S. D.; Webster, K. R. Annu. Rpt. Med. Chem. 2001, 36, 139.
3. (a) Hoehn, H.; Denzel, T.; Janssen, W. J. Heterocycl. Chem. 1972, 9, 235. (b) Denzel, T.; Hoehn, H. Arch. Pharm 1976, 309, 486.
4. (a) Presented in part: Misra, R. N.; Rawlins, D. B.; Bursu- ker, I.; Kellar, K. A.; Kimball, S. D.; Mulheron, J. G.; Sack, J. S.; Shan, W.; Xiao, H. Y.; Webster, K. R. Abstracts of Papers, 219th National Meeting of the American Chemical Society, San Francisco, CA, Mar 26–30, 2000; MEDI038. (b) Misra,
R. N.; Rawlins, D. B.; Batorsky, R.; Bursuker, I.; Kellar, K. A.; Kimball, S. D.; Mulheron, J. G.; Sack, J. S.; Shan, W.; Xiao, H. Y.; Webster, K. R. Abstracts of Papers, 91st Amer- ican Association of Cancer Research National Meeting, San Francisco, CA, April 1–5, 2000; Abstract No. 3002. (c) Misra,
R. N.; Kimball, S. D.; Rawlins, D. B.; Webster, K. R.; Bur- suker, I. US Patent 6,107,305, 2000. (d) Misra, R. N. US Patent 6,448,264 B2, 2002.
5. Misra, R. N.; Rawlins, D. B.; Xiao, H. Y.; Shan, W.; Bur- suker, I.; Kellar, K. A.; Mulheron, J. G.; Sack, J. S.; Tokarski,
J. S.; Kimball, S. D.; Webster, K. R. Bioorg. Med. Chem. Lett.
2003, 13, 1133.
6. Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
7. Ketones 10c and 21l were prepared by addition of benzyl Grignard and 2,4,6-trimethylphenyllithium, respectively, to the acid chloride derived from 7, followed by TFA deprotec- tion.
8. For a description of biological assays and crystallographic procedures see: Kim, K. S.; Kimball, S. D.; Misra, R. N.; Rawlins, D. B.; Hunt, J. T.; Xiao, H.-Y.; Lu, S.; Qian, L.;
Han, W.-C.; Shan, W.; Mitt, T.; Cai, Z.-W.; Poss, M. A.; Zhu,
H.; Sack, J. S.; Tokarski, J. S.; Chang, C.-Y.; Pavletich, N.; Kamath, A.; Humphreys, W. G.; Marathe, P.; Bursuker, I.; Kellar, K. A.; Roongta, U.; Batorsky, R.; Mulheron, J. G.; Bol, D.; Fairchild, C. R.; Lee, F. Y.; Webster, K. R. J. Med. Chem. 2002, 45, 3905.