Toward a Scalable Synthesis and Process for EMA401, Part III: Using an Engineered Phenylalanine Ammonia Lyase Enzyme to Synthesize a Non-natural Phenylalanine Derivative
Leo A. Hardegger, Pascal Beney, Dominique Bixel, Christian Fleury, Feng Gao, Alexandre Grand‐Guillaume Perrenoud, Xingxian Gu, Julien Haber, Tao Hong, Roger Humair, Andreas Kaegi, Michael Kibiger, Florian Kleinbeck, Van Tong Luu, Lukas Padeste, Florian A. Rampf, Thomas Ruch, Thierry Schlama, Eric Sidler, Anikó Udvarhelyi, Bernhard Wietfeld, Yao Yang
Abstract
A process using an engineered phenylalanine ammonia lyase (PAL) enzyme was developed as part of an alternative route to a key intermediate of olodanrigan (EMA401). In the first part of this report, the detailed results from a screening for the optimal reaction conditions are presented, followed by a discussion of several workup strategies investigated. In the PAL-catalyzed reaction, 70–80% conversion of a cinnamic acid derivative to the corresponding phenylalanine derivative could be achieved. The phenylalanine derivative was subsequently telescoped to a Pictet–Spengler reaction with formaldehyde, and the corresponding tetrahydroisoquinoline derivative was isolated in 60–70% yield with >99.9:0.1 er. On the basis of our screenings, carbonate/carbamate-buffered ammonia at an NH3 concentration of 9–10 M and pH 9.5–10.5 was found to be optimal. Enzyme loadings down to 2.5 wt % (E:S = 1:40 w/w) could be achieved, and substrate concentrations between 3–9 v/w (1.17–0.39 M) were found to be compatible with the reaction conditions. A temperature gradient was applied in the final process: a pre-equilibrium was established at 45 °C, before making use of the temperature dependence of the entropy term with subsequent cooling to 20 °C to achieve maximum conversion. This temperature gradient also allowed balancing of the enzyme stability (low at 45 °C, high at 20 °C) with the activity (high at 45 °C, low at 20 °C) in order to achieve optimal conversion (low at 45 °C, high at 20 °C). From the various workup operations investigated, a sequence consisting of denaturation of the enzyme, NH3/CO2 removal by distillation, acidification, and telescoping to the subsequent Pictet–Spengler cyclization was our preferred approach. The process presented in this study is a more sustainable, shorter, and more cost-effective alternative to the previous process.