Late-Stage Amination of Drug-Like Benzoic Acids: Access to Anilines and Drug Conjugates through Directed Iridium-Catalyzed C-H Activation

Abstract

The functionalization of C-H bonds, ubiquitous in drugs and drug-like molecules, represents an important synthetic strategy with the potential to streamline the drug-discovery process. Late-stage aromatic C-N bond-forming reactions are highly desirable, but despite their significance, accessing aminated analogues through direct and selective amination of C-H bonds remains a challenging goal. The method presented herein enables the amination of a wide array of benzoic acids with high selectivity. The robustness of the system is manifested by the large number of functional groups tolerated, which allowed the amination of a diverse array of marketed drugs and drug-like molecules. Furthermore, the introduction of a synthetic handle enabled expeditious access to targeted drug-delivery conjugates, PROTACs, and probes for chemical biology. This rapid access to valuable analogues, combined with operational simplicity and applicability to high-throughput experimentation has the potential to aid and considerably accelerate drug discovery.

Keywords: C−H activation; amination; conjugation; high-throughput experimentation; iridium.

Figure 1

Approaches to late‐stage amination of the Tranilast model system. A) Electrophilic and radical substitutions. Regioselectivity is governed by the steric and electronic properties of the substrate. Pink dots show anticipated reaction sites. S_EAr: electrophilic aromatic substitution; X': nitro, halide. B) Directed C−H activation. Functionalization of C−H bonds in proximal to Lewis basic directing groups. Orange dots show anticipated reaction sites; FG: functional group. C) This work. Carboxylate‐directed regioselective C−H activation. The high regioselectivity is enabled by the transformations proceeding via an iridacycle as depicted.

Scheme 1

C−H sulfonamidation scope. Isolated yields shown. A) Scope of building blocks. [a] TsN₃ (2.1 equiv.) was used. B) Scope of LSF. [b] TsN₃ (2.1 equiv.), KOAc (1.0 equiv.), [Cp*Ir(H₂O)₃]SO₄ (6 mol %). Other potential directing groups are highlighted in purple.

Scheme 2

Additional investigations; isolated yields are shown. A) Single substrate reoptimization. a) Lowest catalyst loading at 0.25 mol %. b) Decreased reaction time (2 h) at 3.0 mol % catalyst loading. c) Use of commercially available [Cp*IrCl₂]₂ catalyst. B) Decarboxylation studies. Anticipated decarboxylation was achieved for copper‐mediated decarboxylation (left), whereas an unexpected heterocycle was formed when the mixture was heated in 1,2‐dichloroethane (DCE).

Scheme 3

C−H amination; isolated yields are shown. [a] Reaction time in first step 44 h. Other potential directing groups are highlighted in purple.

Scheme 4

Synthesis of drug conjugates. The key intermediate was prepared in a two‐step protocol in 77 % overall yield; isolated yields are shown. The yields and step count depicted are from commercially available Tranilast. Each color represents a structural motif added in a new transformation. For detailed reaction conditions, see the Supporting Information.

Scheme 5

Mechanistic studies. A) C−H activation reversibility. Left: reversibility of C−H activation confirmed in the absence of azide source. Right: C−H activation shown to be irreversible in the presence of an azide source. No deuterium incorporation into the starting material was observed. [a] R=Me, 15‐min reaction time; R=F, 8‐min reaction time. B) KIE studies. Left: Parallel experiments with H and D compounds, initial rates measured (see the Supporting Information), TsN₃ as azide source. R=Me k_H/k_D of 2.9 measured. R=F k_H/k_D of 6.2 measured. Right: NsN₃ as azide source. k_H/k_D of 5.5 measured.