Jeffrey G. Varnes a,⇑, Thomas McGuire b,⇑, Rebecca E. Meadows c, Bernard Barlaam b, Jemma Clark b,Calum R. Cook b, Gemma Davison b, Allan Dishington b, Chris De Savi a, Craig Donald b, Tyler Grebe a,Sudhir Hande a, Janet Hawkins b, Alexander W. Hird a, Jane Holmes b, Andrew Lister b, Simon Lucas b,Jane Moore b, Esther Moore b, Anil Patel b, Kurt G. Pike b, Bryan Roberts b, Andrew Stark c, Darren Stead b,Kumar Thakur a, Paul Turner b, Melissa Vasbinder a, Bin Yang a
a Oncology Chemistry, Innovative Medicines Unit, AstraZeneca, 35 Gatehouse Drive, Waltham, MA 02451, USA
b Oncology Chemistry, Innovative Medicines Unit, AstraZeneca, Alderley Park, Macclesfield, Cheshire SK10 4TG, UK
c Chemical Development, AstraZeneca, Silk Road Business Park, Charter Way, Macclesfield, Cheshire SK10 2NA, UK
a b s t r a c t
As part of a medicinal chemistry program, we adapted known synthetic methods for the late-stage diversificationof 2,4-substituted 7-azaindoles. The strengths and weaknesses of these strategies are discussed.In the course of this work, three optimized conditions were identified from an iterative catalyst screen forthe conversion of Boc-protected 4-chloro-2-(piperidin-4-yl)-7-azaindole to the corresponding pinacol borate ester. Additionally, a scalable route to previously unreported Boc-protected 4-bromo-2-(piperidin- 4-yl)-7-azaindole and efficient conversion to the corresponding pinacol borate ester in 72% isolated yield are also disclosed.Azaindoles have generated significant interest as both synthetic targets and key structural components of therapeutic agents.1 As indole or purine isosteres,1d these moieties offer opportunities to modulate physical properties through nitrogen incorporation at the 4-, 5-, 6-, or 7-position, and the variety of potential substitution patterns provides significant opportunity for chemical route development.1a–c Of particular therapeutic interest is the use of 7-azaindoles as kinase inhibitors, where the nitrogens of the pyridine and pyrrole rings can act as a hydrogen bond acceptor and donor pair, respectively, in forming interactions with the protein backbone of the ATP binding site (Fig. 1A).1d As part of a medicinal chemistry effort, we became interested in the preparation of 2,4-substituted 7-azaindoles such as 1 and 2 (Fig. 1), which have been reported in the literature as inhibitors of cyclin-dependent kinase 9 (CDK9)2a–d and Bruton’s tyrosine kinase (BTK),2e,f respectively. We were keen to develop and optimize cost-effective, broadly applicable, and scalable methods for rapid late-stage elaboration of C2 and C4 substituents in order to support internal compound demand for structure– activity relationship (SAR) exploration and in vivo testing. Herein we use compound 1 and related analogs to describe these efforts. Our initial route3 to prepare 1 and related analogs is described in Scheme 1. Starting from bromide 3, which was either purchased from a commercial supplier or prepared from 4-bromo-7-azaindole and tosyl chloride (NaH, DMF; yield: 96%),4 lithiation with LDA (generated in situ) followed by an iodine quench5 afforded 4 in excellent yield. Suzuki-Miyaura coupling conditions using Pd (PPh3)4 preferentially afforded desired carbamate 5,2a while higher temperatures, long reaction times, or more reactive catalysts (e.g., XPhos 2nd Generation Precatalyst6) generally eroded C2 selectivity and drove concomitant oxidative addition at C4. A second Suzuki coupling with 4-fluoro-2-methoxyphenyl boronic acid and XPhos 2nd Generation Precatalyst smoothly provided 6, and the tosyl protecting group was then removed under basic conditions. Yields for deprotection were variable, however, improved yields could generally be obtained using either (a) anhydrous conditions and powdered sodium hydroxide or
(b) aqueous NaOH or KOH with careful reaction monitoring so as to avoid protracted reaction times. Hydrogenation of 7 was accomplished at elevated temperatures to provide piperidine 8, and the Boc protecting group could be removed under standard conditions (TFA/DCM or 4 N HCl in dioxane) to afford 1. Good yields for conversion of 7 to 8 required a hot methanol wash of the catalyst to ensure full recovery, and these conditions are similar to those previously disclosed.2a,c More elaborate scaffolds containing similar olefins, however, have been hydrogenated successfully at room temperature.7 Where a specific C4 substituent was desired (e.g., Ar = 4-F-2- OMe-Ph), the preparation of intermediates such as 10 promoted rapid late-stage C2 diversification under Pd-mediated conditions. We also investigated deprotonation of 9 and addition to electrophiles other than iodine. Such work has been previously reported2a, c,e,f,3,8 for similar substituted and unsubstituted 7-azaindoles, but only one example existed at the time of this writing where a phenyl group was exemplified2a,c at C4. Yields were typically modest (_630%) for successful additions when using lithium diisopropylamide (LDA) as base, and this was attributed to competing ketone alpha-deprotonation, as evidenced by significant amounts of recovered starting material. For 11b, salts such as LaCl3 facilitated addition, and this has been exemplified previously using CeCl3.8a Generally we were satisfied with the route depicted in Scheme 1, as it allowed customization of C2 and C4 substituents. However, this route suffered from being lengthy (6 steps with deprotection) and did not accommodate C4 exploration as the final step. Additionally, we were unable to readily incorporate hydrogenation- labile functionalities. Groups susceptible to strong base were also disadvantaged, and, as a result, we sought a more concise synthesis that offered multiple points of diversification while also being more functionally tolerant. For design hypotheses revolving around specific C2-substituents such as piperidin-4-yl, we adopted an efficient two-step intramolecular cyclization strategy9 similar to that previously reported2a–d,8a (Scheme 2; Ar = 4-F-2-OMe-Ph), which enabled delineation of C4 scope as the penultimate step. Hence, Sonogashira coupling between a commercially available alkyne
(12) and halide (13) followed by cyclization under basic conditions afforded first 14 and then 15 in high yields. In contrast to precedent work, we did not explore the possibility of executing these reactions as part of a one-pot procedure.10 Alternatively, 12 and related alkynes were prepared via Gilbert-Seyferth homologation using the Ohiro-Bestmann reagent.11 Suzuki– Miyaura coupling then afforded 8, while subsequent Miyaura borylation12 using either Pd(dppf)Cl2 or XPhos 2nd Generation Precatalyst failed to deliver appreciable amounts of pinacol borate ester (BPin) 16.13 Based on HPLC analysis, des-chloro 15 was the dominant byproduct in all attempted borylation reactions, and yields of desired product were typically <10% as a mixture of BPin ester and boronic acid (not shown). The route depicted in Scheme 2 overcame many of the challenges inherent in Scheme 1, because it was more functionally tolerant with the absence of hydrogenation and indole deprotectionsteps. This resulted in a shortened overall reaction sequence.A drawback was the early introduction of the C2 moiety, and alkynes which could not be purchased required a custom synthesis from related aldehyde or alcohol precursors. Additionally, a consequence of unsuccessful borylation was that we were unable to easily capitalize on commercial aryl halides as C4 Suzuki–Miyaura coupling partners without additional functional group manipulation. We undertook two parallel approaches to develop improved methods to prepare 16. The first was a two-tier catalyst screening approach for the borylation of 15, which first varied catalyst, ligand, solvent, and base, and analyzed reaction progress after 24 h by HPLC.14 The best conditions for the initial iteration (see Supplemental) based on yield and relative levels of des-chloro byproduct (Ar-H) were found to be Pd(dppf)Cl2, KOAc, and DMA (yield: 64%). Conditions similar to those in the literature (CyJohn- Phos, Pd(OAc)2, KOAc, dioxane, 90 _C) were also successful (yield: 57%) but gave much higher levels of Ar-H. Potassium acetate was a superior base compared to triethylamine or potassiumphosphate. Polar aprotic solvents also minimized Ar-H formation compared to non-polar solvents such as dioxane, with cyclopentylmethyl ether (CPME) being the exception.
Using these data, a second screening iteration was designed, and results are shown in Table 1. In general, electron-rich bulky monophosphine ligands such as dppf, dippf, and Buchwald ligands gave more favorable reaction profiles. Presumably this was through the promotion of both oxidative addition to the challenging heteroaryl-chloride bond and rapid reductive elimination due to steric bulk, thereby reducing the potential for formation of des-chloro byproduct. Pd2dba3 performed poorly compared to Pd (OAc)2, and this may have resulted from formation of a less active catalytic species arising from dibenzylideneacetone (dba) participation.15 The highest yield was measured using Pd(OAc)2/CyJohnPhos in combination with NaOAc and CPME (yield: 87%). This compared favorably to literature conditions tested in the first screening iteration. Similar yields were also obtained with: (a) Pd(OAc)2/PCyPh2 and KOAc in NMP (yield: 85%) and (b) 10 mol % Pd(dppf)Cl2 and KOAc in NMP (yield: 84%). When testing Pd(OAc)2/CyJohnPhos/ NaOAc in CPME on moderate scale in a fume hood (see Supplemental) rather than a glovebox, 16 was isolated in only 31% yield. However, purification of 16 was not wholly successful under normal phase conditions using combinations of solvents such as methanol in dichloromethane or ethyl acetate, and the final product was contaminated with 16 wt % Ar-H. Preparative reverse phase purification using acetonitrile in water containing either 0.1% formic acid or 0.1% ammonium hydroxide was not attempted due to evidence of partial hydrolysis to the corresponding boronic acid under analytical UPLC conditions. While crude mixtures containing Ar-BPin contaminated with small amounts of Ar-H or boronic acid could be used in subsequent Suzuki couplings without further purification, we preferred to develop alternative methods to access 16. Our second strategy to access 16 was predicated on the use of a more reactive bromide. It should be noted that the 4-bromo-3-iodo analog of 13 was not pursued because of (a) a lack of commercial availability and (b) a view that additional steps to prepare such a compound would make the synthesis overlong. Miyaura borylation of a bromide such as 5 had been previously reported,2a–c however, the resulting product contained both undesired olefin and tosyl groups. We therefore first removed the tosyl-protecting group of 5 to afford 17.2e,f Deprotonation with sodium hydride was then followed by transmetalation and lithium–halogen exchange with n-butyllithium; quenching with isopropyl pinacol borate afforded 182b cleanly and in good yield.16 Using sodium hydride to first remove the azaindole NH allowed for controlled lithium–halogenexchange and avoided self-quenching. Gratifyingly, hydrogenation of 18 using atmospheric hydrogen at 55 _C afforded 16 in quantitative yield. We additionally explored reduction of 17 using a ThalesNano Nanotechnology, Inc., H-Cube_ and identified conditions (see Supplemental) which led to olefin reduction in the presence of the C4 bromide to provide 19 in approximately 30% yield based on NMR analysis of the crude reaction mixture. While this transformation is notable because both olefin and bromide are not typically considered to be orthogonal, we were unable to improve this yield, and, as anticipated, desbromo 17 and 19 were major byproducts. Concurrently we developed a more accessible two-step oxidation–halogenation strategy for preparing 4-bromo-2-(piperidin- 4-yl)-7-azaindoles (Scheme 4) based on precedented work
for generating C4-bromides from 7-azaindole-7-oxides.8a,17 We were attracted to this methodology because it allowed us to use the concise cyclization chemistry of Scheme 2 while avoidingindole deprotection and olefin reduction steps. In short, Boc piperidine 21 was prepared starting from 2-amino-3- iodopyridine 20 (_$25/g; Manchester Organics, cat# R10961) and alkyne 12 (_$45/g; Melrob, CAS#104839-06-0) in good yield over two steps (53–61%) on 47 g scale. Oxidation with mCPBA then smoothly afforded N-oxide 22, which was brominated using ammonium bromide in the presence of methanesulfonic anhydride at _5 _C to provide 19. Overall yields for this four-step sequence were 35–40%, and bromide 19 was then smoothly converted to 16 using the aforementioned conditions. To our knowledge, this is the first time 18 has been prepared directly from 17 rather than chloride 162b and the first borylation of 17 without protection of the indole nitrogen.2a–c Additionally, hydrogenations of 17 and 18 to afford 19 and 16, respectively, are also unprecedented, as prior work has either necessitated an unsaturated piperidine or involved reaction sequences similar to Scheme 2 where an olefin was never present.2,3 Borylation conditions of Schemes 3 and 4 were also robust, and Scheme 4 is the first reported syntheses of bromide 19.18 In summary, we adapted precedented chemistries for the late stage diversification of C2 and C4 positions of 2,4-disubstituted 7-azaindoles. Where C2 diversification was desired, N-tosyl protected iodide intermediates such as 10 enabled rapid exploration via cross-coupling or deprotonation chemistries. An alternative two-step cyclization strategy which fixed C2 substitution early on enabled C4 diversification from an aryl chloride as the final step. While shorter, this method failed in our hands to realize adequate quantities of pure BPin 16 from choride 15, limiting our ability to access halogen-containing coupling partners for SAR exploration. A two-tiered catalyst screening approach identified improved methods for generating 16 with minimal formation of des-chloro byproduct. Highest yields were obtained using CPME, which is a green alternative to solvents such as dioxane, THF, and MTBE. We also adapted conditions for preparing 16 via sequential borylation and hydrogenation of known bromide 17. Building on this work, we implemented a two-step oxidation/bromination methodology to prepare bromide 19 on scale in 35–40% yield, and borylation of this material under basic conditions also afforded desired 16 in good yield. As a whole, this work contributes to the synthesis of 4-substituted 2-(piperidin-4-yl)-7-azaindoles by expanding available methods for Miyaura borylation of chloride 15 and disclosing the preparation of a versatile bromide alternative. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.tetlet.2016.09.029. References and notes
1. For recent and leading references, see the following and references therein: (a)
Nuhant, P.; Allais, C.; Chen, M. Z.; Coe, J. W.; Dermenci, A.; Olugbeminiyi, F. O.;
Flick, A. C.; Mousseau, J. J. Org. Lett. 2015, 17, 4292–4295; (b) Kazzouli, S. E.;
Koubachi, J.; Brahmi, N. E.; Guillaumet, G. RSC Adv. 2015, 5, 15292–15397; (c)
Merour, J.-Y.; Routier, S.; Suzenet, F.; Joseph, B. Tetrahedron 2013, 69, 4767–
4834; (d) Merour, J.-Y.; Buron, F.; Ple, K.; Bonnet, P.; Routier, S. Molecules 2014,
19, 19935–19979.
2. Tong, Y.; Bruncko, M.; Clark, R. F.; Curtin, M. L.; Florjancic, A. S.; Frey, R. R.;
Gong, J.; Hansen, T. M.; Ji, Z.; Lai, C.; Mastracchio, A.; Michaelides, M.;
Miyashiro, J.; Risi, R. M.; Song, X.; Tao, Z.-F.; Woods, K. W.; Zhu, G.; Penning, T.;
Souers, A.; Goswami, R.; Iquturi, O. R.; Dabbeeru, M. B. PCT Int. Appl. WO
2014139328 A1, September 18, 2014.; (b) Lai, C.; Tao, Z.-F.; Woods, K. W.;
Penning, T. D.; Souers, A. J.; Mastracchio, A.; Miyashiro, J. M.; Tong, Y. US
20140275153 A1, September 18, 2014.; (c) Gong, J.; Tao, Z.-F.; Tong, Y.; Zhu, G.;
Penning, T.; Souers, A. J. PCT Int. Appl. WO 2014151444 A1, September 25,
2014.; (d) Frey, R.; Gong, J.; Ji, Z.; Lai, C.; Penning, T.; Song, X.; Souers, A.; Tong,
Y.; Zhu, G.-D. US Patent Appl. Pub. US 20150218165 A1, August 6, 2015.; (e)
Zhao, X.; Huang, W.; Wang, Y.; Xin, M.; Jin, Q.; Cai, J.; Tang, F.; Zhao, Y.; Xiang,
H. Bioorg. Med. Chem. 2015, 23, 4344–4353; (f) Jin, Q.; Huang, W.; Zhao, X.;
Wang, Y.; Cai, J.; Yang, J.; Tang, F.; Zhao, Y.; Zhu, J. CN 104211703 A, December
17, 2014.
3. For similar precedented syntheses, see Refs. 2a–c,e,f and the following: Dhanak,
D.; Newlander, K. A. U.S. Pat. Appl. Publ., 20070149561 A1, June 28, 2007.
4. Ledeboer, M. W.; Wannamaker, M. W.; Farmer, L. J.; Wang, T.; Pierce, A. C.;
Martinez-Botella, G.; Bethiel, R. S.; Bemis, G. W.; Wang, J.; Salituro, F. G.;
Arnost, M. J.; Come, J. H.; Green, J.; Stewart, M.; Marhefka, C. WO 2006127587
A1, November 30, 2006.
5. Bamborough, P.; Barker, M. D.; Campos, S. A.; Cousins, R. P. C.; Faulder, P.;
Hobbs, H.; Holmes, D. S.; Johnston, M. J.; Liddle, J.; Payne, J. J.; Pritchard, J. M.;
Whitworth, C. WO 2008034860 A1, March 27, 2008.
6. Kinzel, T.; Zhang, Y.; Buchwald, S. L. J. Am. Chem. Soc. 2010, 132, 14073–14075.
7. (a) Ahrendt, K. A.; Buckmelter, A. J.; Grina, J.; Hansen, J. D.; Laird, E. R.; Moreno,
D.; Newhouse, B.; Ren, L.; Wenglowsky, S. M.; Bainian, F.; Gunzner, J.; Malesky,
K.; Mathieu, S.; Rudolph, J.; Wen, Z.; Young, W. B.; PCT Int. Appl. WO
2009111278 A2, September 11, 2009.; (b) Wu, L.; Zhang, C.; He, C.; Sun, Y.; Lu,
L.; Qian, D.-Q.; Xu, M.; Zhuo, J.; Yao, W. U.S. Pat. Appl. Pub. 20130338134,
December 19 2013.
8. For representative examples of C2 deprotonation and addition to electrophiles
for 7-azaindoles, see the following: (a) Stoit, A. R.; den Hartog, A. P.; Mons, H.;
van Schaik, S.; Barkhuijsen, N.; Stroomer, C.; Coolen, H. K. A. C.; Reinders, J. H.;
Adolfs, T. J. P.; van der Neut, M.; Keizer, H.; Kruse, C. G. Bioorg. Med. Chem. Lett.
2008, 18, 188–193; (b) Paczal, A.; Balint, B.; Weber, C.; Szabo, Z. B.; Ondi, L.;
Theret, I.; De Ceuninck, F.; Bernard, C.; Ktorza, A.; Perron-Sierra, F.; Kotschy, A.
J. Med. Chem. 2016, 59, 687–706.
9. (a) Koradin, C.; Dohle, W.; Rodriguez, A. L.; Schmid, B.; Knochel, P. Tetrahedron
2003, 59, 1571–1587; (b) Schirock, H. J. Org. Chem. 2006, 71, 5538–5545.
10. McLaughlin, M.; Palucki, M.; Davies, I. W. Org. Lett. 2006, 8, 3307–3310.
11. For respresentative Gilbert-Seyferth reactons see: Arkin, M. R.; McDowell, R. S.;
Oslob, J. D.; Raimundo, B. C.; Waal, N. D.; Yu, C. H. PCT Int. Appl. WO
2003051797 A2, June 26, 2003.
12. Ishiyama, T.; Murata, M.; Miyaura, N. J. Org. Chem. 1995, 60, 7508–7510.
13. Since the time of this work, conditions have been reported for the conversion
of 15 to 16 using CyJohnPhos and Pd(OAc)2 in dioxane but yields were not
reported. See Refs. 2a,c.
14. See Supplmental for full details on both initial and follow-up screening results,
as well as a Table depicting %Ar-H relative to an internal standard for each
reaction in Table 1.
15. Amatore, C.; Jutand, A.; Khalil, F.; M’Barki, M. A.; Mottier, L. Organometallics
1993, 12, 3168–3178.
16. For references detailing the preparation of C4 boronic acids from C4 bromides
of unsubstituted indole and 7-azaindole using NaH and nBuLi, see Ref. 5 and:
Noronha, G.; Barrett, K.; Cao, J.; Gritzen, C.; Gong, X.; Hood, J.; Mak, C. C.;
Mcpherson, A.; Pathak, V. P.; Renick, J.; Soll, R.; Splittgerber, U.; Wrasidlo, W.;
Zeng, B.; Zhao, N.; Dneprovskaia, E. WO 2005096784 A2, October 20, 2005.
17. For representative oxidation-bromination of C2-substituted 7-azaindoles in
excellent yield, see: (a) Adams, N. D.; Adams, J. L.; Burgess, J. L.; Chaudhari, A.
M.; Copeland, R. A.; Donatelli, C. A.; Drewry, D. H.; Fisher, K. E.; Toshihiro, H.;
Hardwicke, M. A.; Huffman, W. F.; Koretke-Brown, K. K.; Lai, Z. V.; McDonald, O.
B.; Nakamura, H.; Newlander, K. A.; Oleykowski, C. A.; Parrish, C. A.; Patrick, D.
R.; Plant, R.; Sarpong, M. A.; Sasaki, K.; Schmidt, S. J.; Silva, D. J.; Sutton, D.;
Tang, J.; Thompson, C. S.; Tummino, P. J.; Wang, J. C.; Xiang, H.; Yang, J.;
Dhanak, D. J. Med. Chem. 2010, 53, 3973–4001.
18. For synthesis of the 4-bromo-5-chloro analog of 19, which starts from
commercially available 4-bromo-5-chloro-3-iodo 2-aminopyridine and uses a
method analogous to that of Scheme 2, see Refs. 2a,c.
