Biocatalysis is widely regarded as a sustainable technology to innovate Active Pharmaceutical Ingredient (API) synthesis. Biotransformation is considered a unique technology in the drug manufacturing world to unlock synthetic routes to complex chiral motifs. Sanofi is accelerating the development of strong mindset and efficient capabilities in biocatalysis to foster API delivery and promote greener solutions for its synthetic processes.In this article, we will describe how Sanofi is building its internal biocatalytic capabilities in order to address new challenges related to sustainability and the growing complexity of its portfolio. The outcomes of recent studies will be presented demonstrating the successful implementation of the biocatalysis technology across early and late-stage development of our pipeline assets.

Sotorasib is a first-in-class KRAS(G12C) inhibitor with a unique carbon-nitrogen atropisomer. Described herein is the development of an improved second-generation process to manufacture a key sotorasib intermediate, rac- 4. Notable improvements to the second-generation process include the development of a three-step fully telescoped sequence in acetonitrile and the use of phosgene as an isocyanating agent, which allowed for dichloromethane, aqueous extractive work-ups, and distillations to be completely removed from the process. Compared to the first-generation process, these process improvements led to a 97% reduction in organic solvent usage and a 71% reduction in wastewater generation and still maintains a 75% yield and >99.5% purity of rac-4 in the overall process.

Herein, a novel route to atropisomeric N-arylquinolones with low rotational barriers is demonstrated, leveraginga dual photochemical/organocatalytic approach to the required ringclosure in up to 94% yield and up to >99% ee. The use of a continuousflow system allows for impurity suppression and enables rapid scale-upto a decagram scale.

A chiral Cu(II)-PyBOX complex was prepared and utilized to accomplish desymmetrization of a 1,3-diol en route to AMG 176. An expedient synthesis of the racemic 1,3-diol is described, in addition to efforts for rapid optimization and subsequent kilogram-scale execution of the desymmetrization process to generate material for pre-clinical activities. An overview of the kilogram-scale synthesis of the Cu(II)-PyBOX complex is provided.

A commercial process to manufacture sotorasib (AMG 510), a first-in-class KRASG12C inhibitor, is described. Development efforts focused on rendering a fit-for-purpose early-phase route into a viable long-term commercial process through the reduction of side reactions to improve yield and product quality, as well as reducing cycle times of crystallization processes by improving particle properties and filtration times. These improvements were key to ensuring clinical supply and commercial launch. The final route consists of five synthetic operations from starting material M-1, including a telescoped two-step sequence, and a final form-setting crystallization.

A three-step synthesis of (S)-TRIP enabled by efficient Suzuki cross-coupling conditions using commercial starting materials was developed and demonstrated on a kilogram scale. These novel Suzuki reaction conditions feature Pd-2(dba)(3)/CataCXium A in the presence of TBAB and KOH and provide conversions up to 90% while minimizing the formation of common byproducts. Following an improved demethylation protocol and a powerful methanol purification protocol during step 2, high-quality catalyst of up to 99% purity was isolated in 52% yield over three steps.

Anilines are valuable synthons in pharmaceuticals and agrochemicals. These compounds are generally produced by chemocatalytic reduction of the corresponding nitrobenzene precursors. However, known synthetic methods often lack sufficient activity or selectivity, which results in low yields or the formation of a variety of undesired side products. We envisaged a biocatalytic approach as a promising general platform for selective and mild nitroarene reduction. Herein, we report using nitroreductases in combination with vanadium salts for the quantitative reduction of nitroaromatics to their corresponding anilines. Substrate scope studies were performed with fourteen nitrobenzene and four nitropyridine compounds. In one example, the reaction was intensified to 27 ​g/L substrate loading at 25 ​mL scale, where chemoselective reduction of the nitro group was obtained with full conversion and more than 93% selectivity toward aniline product (isolated in 82% yield). These conditions demonstrate the first general enzymatic method for the reduction of nitroaromatics to anilines.

AMG 176 is a drug candidate in our oncology pipeline. Ring-closing metathesis (RCM) is a key reaction in the AMG 176 synthesis that is used to construct the 16-membered macrocycle. Process intensification was executed on a compressed timeline by combining high-throughput experimentation and computational analysis using density functional theory, which led to the identification of an optimal 4-bromobenzoate protecting group on the allylic alcohol moiety. Comprehensive process improvements led to a reduction in reaction volume from 800 to 50 L/kg with a >20% yield improvement compared with the discovery route. The RCM process developed from these studies was instrumental in the clinical advancement of the AMG 176 program.

Reduction of aromatic nitro compounds to anilines is of great interest to the chemical industry. Biocatalytic reduction of nitroarenes has made it possible to effectively produce anilines by applying nitroreductase enzymes (NR) in combination with vanadium pentoxide. Herein, the NR-catalyzed reduction of 2-methyl-5-nitro-pyridine (2) to give the desired aniline (1) was studied as a model reaction. It demonstrates the importance of process development and enzyme engineering as key approaches to overcome scale-up issues and improve yield and productivity. Moving to fed-batch allowed controlling the feeding rate of 2 to prevent the accumulation of intermediates and formation of undesired side products. Starting with a substrate (2) concentration of 200 mM (28 g/L) and enzyme loading of 5 mg/mL (18% w/w), it was possible to achieve complete conversion and 1 in 95% yield by high-performance liquid chromatography (89.1% isolated yield) over 18 h, whereas, with 500 mM (69 g/L) 2 and an enzyme loading of 10 mg/mL (14.5% w/w), the same conversion and yield were achieved in 26 h. A rational engineering of NR-4 yielded faster variants, including NR-5, in only one round. The improved rate of the new variants allowed increasing the feeding rate of 2 to shorten the reaction time to less than a day as well as decreasing the enzyme loading to 3.6%.

The last decade of small-molecule process development has witnessed a trend of increasing molecular complexity for clinical candidates. The continued advance of novel catalytic methods and subsequent translation to efficient and scalable processes has enabled process chemists to overcome the challenges associated with increasing complexity. This Account highlights several examples from the process chemistry laboratories at Amgen. 1 Introduction 2 The Evolution of Molecular Complexity 3 Catalysis as a Lever to Build Complexity 4 Ru(II)-Catalyzed Dynamic Kinetic Resolution Enabling the Manufacture of AMG 232 5 Application of Enzymatic Desymmetrization toward Scale-Up of the MCL-1 Inhibitor AMG 176 6 Synthesis of Fucostatin 1: Catalytic Asymmetric Transfer Hydrogenation 7 Manganese-Catalyzed Asymmetric Epoxidation To Prepare a Carfilzomib Intermediate 8 Asymmetric Reduction Strategies: Novel Apelin Receptor Agonists and AMG 986 9 Conclusions

We describe herein the development of a scalable Noyori reductive dynamic kinetic resolution to manufacture DLAC, a Delta-lactone precursor to the active pharmaceutical ingredient AMG 232. Central to this work was the identification of the ruthenabicyclic complex RuCl[(S)-daipena][(S)-xyIBINAP] ((S)-RUCY-xyIBINAP), which afforded the product with >98:2 enantiomeric ratio at a substrate to catalyst loading (S/C) of 2000:1. By transesterification to a more sterically hindered isopropyl ester prior to the hydrogenation, we were able to curb unexpected ester reduction. Optimization of base equivalents in the final alkylation step to form DLAC prevented product degradation. The optimized process was scaled to >200 kg, providing 147 kg of DLAC in 56% overall yield with 99.9% optical purity.

Monitoring chemical reactions by nuclear magnetic resonance (NMR) is an established and valuable approach for process understanding, robustness, scalability, and control in the pharmaceutical industry. Understanding speciation, reaction rates, and reaction completion times provides information on how to improve a chemical process, leading to increased quality and quantity of the desired product. An important consideration for online monitoring is to have an NMR instrument colocated with a chemical reactor. The standard commercial medium- to high-field NMR instruments are normally installed in isolated locations due to facility and safety restrictions. Low-field NMR instruments suffer from low resolution and sensitivity, requiring chemometric analysis for medium to complex chemical structures. Reactions are typically monitored using NMR tubes and deuterated solvents. Therefore, reaction analysis may not provide the same kinetic information as when the reaction occurs in a reactor at a larger scale. To overcome these factors, we have tested a prototype NMR instrument with a 400 MHz cryogen-free power-driven high-temperature superconducting (HTS) magnet installed in a chemistry laboratory fume hood for online monitoring of reactions. We have tested the HTS NMR system with a ring-closing metathesis (RCM) reaction of diethyl diallyl malonate with Grubbs 2nd Gen catalyst in a reactor with a protonated solvent. The reaction was monitored online with a Bruker InsightMR flow cell, and data was acquired in automation, yielding a kinetic time-course of the transformation and reaction rate values. This work demonstrates that NMR instruments with HTS magnets can be integrated into the chemistry laboratory with other equipment and are a valuable tool for reaction monitoring under typical reaction conditions and in protonated solvents.

This article details the approach to large-scale production of cyclobutane 2 by the continuous-flow [2 + 2] photocycloaddition of maleic anhydride and ethylene, including (1) focused reaction optimization and development of a robust isolation protocol, (2) the approach to equipment design and process safety, and (3) the results of commissioning tests and production runs delivering the target compound at throughputs exceeding 5 kg/day.

The development of a phase-appropriate manufacturing-scale synthesis of potassium 2-fluoro-6-hydroxyphenyltrifluoroborate was achieved. Investigations into improving the yield and robustness indicated that pH of the reaction medium is a critical process parameter. Additional development resulted in replacing tartaric acid with citric acid, resulting in improved process robustness and enabling scale-up to >10 kg. (C) 2019 Elsevier Ltd. All rights reserved.

A Perspective of our work in the development of innovative synthetic methods within the discipline of Process Research and Development is presented. Through an overview of some of the programs that we have worked on during the past decade, we have selected cases studies to illustrate the challenges faced in development of robust chemical processes for molecules on a multi kilogram scale. The examples have been selected to demonstrate the innovative chemistry being developed within our laboratories with a focus on fragment design, asymmetric synthesis, new synthetic reagents, and the methods that have allowed us to deliver cost-effective syntheses under reduced timelines in an increasingly competitive environment. The technical challenges are presented in the context of molecule complexity that while increasing in the portfolio of small molecules being developed inspires us to deliver new solutions. Overall, our goal is to highlight the exciting work that can be done within our field to support the discovery and delivery of medicines to patients.

A normal phase liquid chromatography method was developed for the separation and detection of eight stereoisomers of the key intermediate, CORE + OMe, having three chiral centers. The stereochemistry of this intermediate dictates the stereochemistry of the active pharmaceutical ingredient generated by an additional six synthetic steps. Multiple columns and mobile phases were screened during the development based on a platform approach. The use of dichloromethane as mobile phase additive and adjustment of flow rate and column temperature contributed in achieving resolution of these eight stereoisomers. The separation and detection of these stereoisomers was achieved using a Chiralcel OD-H, 4.6 x 250 mm, 5 mu m d(p) column with heptane: ethanol: dichloromethane in a ratio of 95:3:2 (v:v:v) as mobile phase at a flow rate of 0.7 mL/min. UV detection was carried out at 245 nm and the column temperature was maintained at 15 degrees C. The analytical method was phase appropriately validated. The limit of detection and limit of quantification were found to be 0.035 and 0.0714 mu g, respectively. The newly developed method has been implemented for routine utilization to monitor the chiral control during process development and used as the quality control method for chiral purity of the desired compound. (C) 2018 Elsevier B.V. All rights reserved.

We present the application of the common bidentate phosphine ligand Xantphos toward the highly selective Negishi cross-coupling of heteroaryl halides and acyclic sec-alkyl organozinc reagents to prepare pharmaceutically relevant motifs. Branched-to-linear ratios of >100:1 can be achieved for several substrates relevant to the pharmaceutical industry, and tolerance of certain acidic protons is exhibited. A high-throughput experimentation approach was taken to rapidly compare Xantphos Pd G3 to other selective Negishi coupling catalysts, leading to separate reactivity profiles for each methodology. The utility of Xantphos Pd G3 was demonstrated through the scale-up and isolation of a complex pyridine building block.

During the formation of a tetrazole ring on an investigational drug, two in-process impurities were detected and analyzed by LC-MS, which suggested that both impurities were drug-related with the same mass-to-charge ratio. To understand and control their formation, both impurities were isolated from the mother liquor of the reaction using a multi-step isolation procedure to obtain a sufficient amount for high-resolution mass spectrometry (HRMS) and NMR structural analysis. HRMS suggested a protonated mass of 577.32 Da for both impurities; however, MS fragmentation patterns provided limited information on their structures. NMR analysis indicated the presence on an additional NH functional group in both isolates with similar spatial and bond correlations to one of the dimethylcarbamoyl moieties and the corresponding aromatic ring. A phenyldimethylcarbamoylamino moiety was supported by the NMR and HRMS data and could be explained based on the 'Schmidt-like' reaction mechanism, which was an unexpected reaction pathway. Because the reaction conditions were fixed because of safety concerns, the crystallization protocol was redesigned to reduce the levels of these impurities significantly. Copyright (C) 2016 John Wiley & Sons, Ltd.

Herein we report the hydroxyethylamine (HEA)-derived potent and orally efficacious BACE1 inhibitors as potential treatments for Alzheimer's disease. These compounds were designed for low efflux and in vivo clearance to effect robust reduction of A beta levels in the central nervous system (CNS). Key design strategies feature an amide masking approach for mitigating PGP-mediated efflux and the incorporation of CYP 3A4 inhibitory activity for decreased in vitro and in vivo clearance. Lead molecules demonstrated sufficient oral bioavailability and CNS penetration and were shown to be orally efficacious in pre-clinical rodent models. Collaboration between medicinal and process chemistry on the key synthetic challenges is presented including new chemistry towards challenging fragments of the HEA core structure. The new routes were designed for scalability and improved overall safety (elimination of hazardous reagents). Additionally a new, templated assembly route toward the HEA core structures was developed to overcome key challenges using traditional methods for HEA construction.