The beginnings. When we started setting up our laboratory at IOCB in Prague we were eager to develop synthesis methods and strategies toward atropisomers, conformational isomers which are becoming increasingly popular in drug discovery[1]. Our main goal was to provide efficient, fast, modular and programmable ways to synthesize libraries of drug-like atropisomers for Med Chem. Two intrinsic features of atropisomers are the most crucial when thinking about and justifying our projects (Figure 1A). First, atropisomers are sterically hindered compounds; thus, presumably, if one develops efficient reactions toward their synthesis, one will be developing efficient reactions per se. Second, atropisomers are three dimensional, currently the most elusive and sought after shape of bioactive molecules employed in early stages of medicinal chemistry screenings[2],[3]. As one can imagine, to work on such topic while establishing the new laboratory came with numerous challenges but as the saying goes, “necessity is the mother of invention”. While we were purchasing and organizing chemicals and instruments, we came across Nakamura’s magnificent synthesis of Cihunamide B[4], an intriguing and naturally-occurring antibacterial atropisomeric macrocyclic peptide (Figure 1B). The first step in the synthesis was an efficient nucleophilic aromatic substitution (SNAr) involving a Boc-protected tryptophan with a free carboxylic acid (!) unit and an active aryl fluoride. A literature search revealed such reactions, at least for indole and carbazole derivatives, were efficient for synthesis of relatively unhindered C─N bonds, but would SNAr allow for efficient and broadly applicable synthesis of C─N atropisomers? Answering this question led us to an enjoyable and somewhat unexpected journey, to the point where it is now perhaps not an exaggeration to consider SNAr as being a click reaction between N-heterocycles and electron-deficient fluoroarenes. We developed practical conditions to synthesize many diversifiable atropisomers which can give rise to an even larger number of drug-like three dimensional compounds. Possibilities are endless, and we expect this work to be employed widely by medicinal chemistry groups. It is amusing but not surprising that inspiration came from a target-oriented approach, underscoring why we will continue to read about and study natural product total synthesis in our group meetings.
Figure 1. Goals, motivations, inspiration and development of a click-type reaction toward C–N atropisomers.
The developments. Our optimized SNAr conditions are accessible to any laboratory in the world (Figure 1C). It takes one equivalent of the N-heterocycle, 1.0 equivalent of the fluoro(or chloro)-arene, 2.0 equiv of Cs2CO3 and reagent grade DMSO in an open container. The scope is extremely broad, and we call the base-promoted process method A (Figure 2A). Of course, pKa considerations are relevant and we have got N─H ranging from 10-21 pKa units to work, and fluoroarenes need to be electron deficient though we show that by increasing temperature even unactivated electrophiles engage productive atropisomer formation. To make it more appealing to medicinal chemists, we validated a high-throughput synthesis approach (Figure 2B) that can be implemented in Direct-to-Biology (D2B) screenings[5]. Why could this be done? Because SNAr reactions are high yielding, produce insoluble salts that can be filtered off as by-product, and are carried out in DMSO, often the solvent of choice in medicinal chemistry screenings. We show herein some of the atropisomers we synthesized, but please note one must go to Supplementary Information to see all that can be done. We provide synthesis of a large number of heterocyclic C─N atropisomers which are diverse and diversifiable!
Figure 2. SNAr provides diverse and diversifiable atropisomers and conditions are amenable to Med Chem.
Of course, we had some problems in scope along the way and, if we were to develop atroposelective versions, perhaps method A is not ideal because controlling background reactivity might be difficult. To address those and other problems, we developed not one but two catalytic SNAr variants (Figure 3): 1) method B employing N─SiR3 heterocycles and TBAT (or TBAF) as catalyst[6], and 2) method C, an improved version employing N─H heterocycles directly, and promoted by Ruppert-Prakash’s reagent (Me3SiCF3) in combination with a fluoride source[7]. Among other advantages, the catalytic methods can lead to faster reactions (in cases complete in seconds) and functional groups that do not tolerate bases such as Cs2CO3 can survive.
Figure 3. Development of two fast catalytic reactions to broaden scope and more.
We were puzzled by the broad scope and efficiency of the reactions, specially because we were dealing with compounds that are notoriously difficult to synthesize. We teamed up with Daniel Bím, a scientist starting his independent career at University of Chemistry and Technology (UCT Prague, 400 m away from us) for DFT calculations. We observed reactions proceed in a stepwise fashion; because the electrophile is activated, a stable Meisenheimer intermediate is formed and the addition step is rate-determining. Nothing new about this. However, a closer look into the Meisenheimer reveals that this intermediate is not atropisomeric because its C─N barrier to rotation is lower than atropisomerism threshold and less than half of that of the product. (Figure 4A). Why does this matter? Because this shows sterically hindered compounds (atropisomers) can be synthesized through transition states that might not be hindered (non-atropisomeric). We hope this idea can help chemists to develop better methods toward synthesis of sterically hindered compounds (Figure 4B).
Figure 4. Atropisomer formation via non-atropisomeric (and unhindered!) intermediates and transition states.
Our work was divided in two phases: 1) the abovementioned assembly by SNAr followed by 2) diversification by myriad methods. The modularity of the approach enabled us to synthesize, readily and rapidly, countless compounds inspired by heteroaryl-containing bioactive molecules: atropisomeric macrocycles including peptides, conformationally modified pharmaceuticals and compounds derived from click and other reactions, highlighting potential for drug discovery (Figure 5).
Figure 5. Diversity-oriented syntheses showcase modularity and programmability of the approach.
The future. The above reactions and concepts have already sparked other projects in our lab and institute, but we hope they get used elsewhere, in academic and industrial environments for synthesis of bioactive compounds with hindered bonds. We are now confident SNAr reactions can tackle those challenges.
Author:
Paulo H. S. Paioti
Group Leader at the Institute of Organic Chemistry and Biochemistry (IOCB Prague)
Contributing authors:
Michal Šimek
Postdoc at IOCB Prague
Paritosh Dey
PhD student at IOCB Prague
Vilém Blahout
PhD student at IOCB Prague
REFERENCES
[1]. Clayden, J., Moran, W. J., Edwards, P. J. & LaPlante, S. R. The challenge of atropisomerism in drug discovery. Angew. Chem., Int. Ed. 48, 6398–6401 (2009).
[2]. Hung, A. W. et al. Route to three-dimensional fragments using diversity-oriented synthesis. Natl. Acad. Sci. U.S.A. 108, 6799−6804 (2011).
[3]. Brown, D. G. & Boström, J. Analysis of past and present synthetic methodologies on medicinal chemistry: where have all the new reactions gone? J. Med. Chem. 59, 4443–4458 (2016).
[4]. Yu, L.; Nagata, Y. & Nakamura, H. Atroposelective synthesis of Cihunamide B. J. Am. Chem. Soc. 146, 2549–2555 (2024).
[5]. Thomas, R. P. et al. A direct-to-biology high-throughput chemistry approach to reactive fragment screening. Chem. Sci. 12, 12098−12106 (2021).
[6]. Sekino, K. et al. Fluoride-ion-catalyzed synthesis of ladder-type conjugated benzobisbenzofurans via intramolecular nucleophilic aromatic substitution reaction under metal-free and mild conditions. Org. Lett. 22, 2892−2896 (2020).
[7]. Liu, X., Xu, C., Wang, M. & Liu, Q. Trifluoromethyltrimethylsilane: nucleophilic trifluoromethylation and beyond. Chem. Rev. 115, 683−730 (2015).