Advancing API Synthesis

Process chemists in the fine-chemicals and pharmaceutical industries are tasked with developing optimal routes for manufacturing pharmaceutical intermediates and APIs. Among their challenges, they must develop approaches to improve yield, purity, stereoselectivity, and solid-state properties for a given API while optimizing production economics as a product moves from development to commercial scale. Some interesting recent developments include commercial-scale amide formation and an improved process route for a tetracycline derivative.

Commercial-scale amide formation
It is well known that amide-formation chemistry can be inefficient and warrants further investigation. This issue has been addressed in the chemical literature, most recently in a study by the American Chemical Society Green Chemistry Institute Roundtable that is particularly relevant to pharmaceutical synthesis (1). The study found that, out of a random selection of drug candidates, amide-bond formation was used in the synthesis of 84% of drug candidates.

The only theoretical by-product of amide formation is water, but examples of this type of reaction are incredibly rare, according to Barrie Rhodes, director of technology development for the CMO Aesica. “Frequently,” he says, “commercial-scale amide syntheses for pharmaceutical manufacture require overly complex stoichiometric coupling agents or reagents.”

Aesica has set as goals the reduction of this complexity in conventional amide syntheses and the development of more sustainable (green) chemical transformations that are practical on a commercial scale. In the pursuit of those goals, the company has partnered with the University of Nottingham for the commercial development of alternative methods in amide-bond synthesis. The partnership’s aim is to revolutionize traditional amide-formation techniques by generating alternative methods for amide-bond formation that will be more eco-friendly and chemically versatile, according to Rhodes.

The new approach should be commercially available to Aesica customers later in 2013. The company is actively seeking commercial opportunities to work with potential compounds that could benefit from the novel technology. “We envisage this new development helping pharmaceutical companies that encounter problems with amide synthesis, and due to the utilization of more sustainable reagents, production costs will be lowered while chemical yields will be increased,” Rhodes notes.

The initial chemistry was developed in 2005 by Simon Woodward, professor of synthetic organic chemistry at the University of Nottingham in the United Kingdom. The coupling reagent of interest is DABAL-Me3, which is an adduct of trimethylaluminum and DABCO (1,4-diazabicyclo[2.2.2]octane). Unlike trimethylaluminum which is very pyrophoric, DABAL-Me3 is a free-flowing solid that can be handled in air (2). In addition to its use in amide-bond formation (3), DABAL-Me3 has been used for the methylation of aldehydes and imines (4, 5), the methylation of aryl and vinyl halides (6), and conjugate additions to enones (7).

With respect to amide bond-formation, DABAL-Me3 can be used to generate amides from unactivated esters and amines that, with conventional routes, require the use of trimethylaluminum or diisobutylaluminum hydride (3). In addition, reactions with DABAL-Me3 tolerate various functional groups, including acetals, alcohols, alkenes, alkynes, ethers, nitriles, hindered esters, and BOC groups. Stereocenters in non-peptidic species are not racemized. Importantly, the preparation of aromatic and aliphatic amides can generally be carried out in an air atmosphere. It should be noted that the rate of the reaction can be accelerated with the use of microwave irradiation, and products can be isolated in 51-99% yield in 8-16 minutes (8).

Preliminary studies on DABAL-Me3 at the university were undertaken using funds awarded by the Engineering and Physical Sciences Research Council (EPSRC) under the Research Development (Pathways to Impact) Funding Scheme. “Since realizing the initial development of our coupling agent in 2005, one of our goals has been to see this novel technology used in larger-scale industrial environments,” remarks Woodward. “We look forward to collaborating with Aesica and seeing the full commercial potential of this novel technology in API manufacture,” he adds.

The chemistry that Aesica is commercializing is more atom-efficient than some other types of amide-formation chemistry and offers a novel synthetic route to make amides from both esters and carboxylic acids, according to Rhodes. Some of the technology is in the very early stages of development and will likely be patentable, so Rhodes is unable to disclose any additional details. He does note that the chemistry is generally applicable and flexible in terms of its ability to prepare amides, and, therefore, any API that either contains amide bonds or goes through an amide intermediate during its synthesis could benefit from this technology. In addition, Rhodes believes that the new amide production technology will enable cheaper and simpler routes to market for many compounds.

This partnership with the University of Nottingham is the Aesica Innovation Board’s (AIB) fourth with an academic institution in less than six months, according to Rhodes. The AIB was established to help bridge the growing R&D gap by identifying early-stage technologies for development into commercial applications.

“The University of Nottingham is renowned for its excellence in chemistry research and has a strong background in green and sustainable chemistry. That, coupled with its interest in open innovation (in that risk and reward are shared) as a model, has been very beneficial. Effectively, the university has the expertise in terms of the technology while Aesica brings its expertise in terms of commercialization and a global network in the pharmaceutical industry,” Rhodes explains.

The partnership for the development of amide bond-formation chemistry is just the start of a hopefully long-term collaboration between Aesica and the university, according to Rhodes. The collaboration builds upon announced plans by the University of Nottingham to establish a Center of Excellence for Sustainable Chemistry, which will be partly funded by an investment from the Higher Education Funding Council for England UK Research Partnership Investment Fund. The Center aims to form creative partnerships with innovative companies to develop new chemical-based technologies that minimize environmental impact and are both energy and resource efficient, according to a university press release.

“As Aesica further enhances its innovation program, we will seek to develop new technologies, not only with the University of Nottingham, but with other academic institutions as well, in the fields of both API and formulated products manufacture,” concludes Rhodes.

Process-scale synthesis of tetracycline derivative
Tetracyclines comprise a group of antibiotics that are recognized as safe and effective and are thus commonly used to treat serious bacterial infections and other less severe conditions such as acne. Unfortunately, because tetracyclines are commonly used, many bacteria have developed resistance to the older versions of these drugs. Recent efforts have thus been directed at developing new tetracycline derivatives.

Scientists at Tetraphase Pharmaceuticals are overcoming this barrier by implementing a new synthetic route first reported by Myers in 2005 (9). This approach involves the coupling of a cyclohexenone intermediate that contains the key tetracycline functionalities with a second functionalized aromatic intermediate via a Michael-Dieckmann reaction, thus enabling the incorporation of a variety of different substituents at various positions in the tetracycline skeleton. Using this methodology, Magnus Ronn, vice-president of CMC at Tetraphase Pharmaceuticals and his colleagues at the company recently reported the successful preparation of eravacycline, a fully synthetic broad spectrum 7-fluorotetracycline in clinical development, in multihundred gram quantities (10). A summary of their work is presented below.

The advantage of this approach to the synthesis of tetracycline analogues is that a single key intermediate can be used to access a wide range of substituted tetracycline active pharmaceutical ingredients (APIs),” says Ronn. This key intermediate is a tricyclic cyclohexenone with three chiral centers  (the synthesis of this compound was reported previously [11]). The enone is reacted with a suitably functionalized phenol bearing an ortho-carboxyphenyl group and a meta-methyl substituent. Other functionalities are included as needed to produce the desired tetracycline analogue.

This aromatic compound, referred to by the researchers as the lefthand piece (LHP), is deprotonated with a strong base to form a benzylic anion, which then undergoes diastereoselective 1,4-conjugate (Michael) addition to the enone moiety when added to the cyclohexenone. The ketone enolate that forms from this step undergoes a Dieckmann-type condensation with the phenyl ester to produce the protected tetracycline compound. To obtain the desired tetracycline analogue, this intermediate is subjected to subsequent silyl-ether cleavage and hydrogenolysis of the benzyl protecting groups with concomitant reductive ring opening of the isoxazole (10). The LHP selected for the preparation of eravacycline is a benzyl-protected phenol with a fluorine atom and a dibenzylamine substituent. It was prepared from a commercially available starting material in seven steps, the synthesis of which will be published in the future (10).

One of the hurdles that the researchers had to overcome in developing the large-scale synthesis of eravacycline was the sensitivity of the Michael−Dieckmann transformation to the reaction conditions, according to Ronn. Not only the order of addition, but the strength of the base was important for the two different deprotonation steps (10). Thus, the researchers reported that it was necessary to first deprotonate the LHP (1.04 equivalents of LHP is used) with lithium diisopropylamide (LDA, 1.13 equivalents) and then add the generated anion to a solution of the cyclohexenone and the weaker base lithium bistrimethylsilylamide (LiHMDS) at -70 °C. The desired adduct was isolated after workup and trituration with methanol in > 90% yield a 98% purity (using high-performance liquid chromatography), even on the 200-g scale (10).

Because both the deprotonation and the Michael−Dieckmann reaction should be performed at -70°C, two cryogenic reactors are required. The researchers reported that attempts to eliminate one of those reactions by raising the temperature of the cyclohexenone solution to -20 °C led to increased production of impurities (10).

To obtain eravacycline, the first step after the Michael-Dieckmann reaction involved cleavage of the tert-butyl silyl (TBS) protecting group. Despite the issues associated with using hydrofluoric acid in commercial manufacturing, the researchers reported that this reagent gave better results than other investigated alternatives and it was thus selected for scale-up (10).

Reductive ring opening of the isoxazoline group and removal of the four benzyl groups using palladium on carbon(Pd/C)/hydrogen to give the 9-amino-7-fluoro-sancycline required extensive investigation by the researchers (10). A mixed solvent system of tetrahydrofuran (THF) in methanol (1:3) was required because of solubility issues. An acid additive was also needed to improve the rate of the hydrogenation reaction, but epimerization at the C-4 position and reduction of undesired groups led to the formation of impurities, including one that was very difficult to separate from the desired product. The reaction was optimized using concentrated aqueous hydrochloric acid (HCl) because it is a stable reagent with a reliable concentration. The palladium on carbon was removed using Celite, and residual palladium was eliminated with the metal scavenger (SiliaBond DMT, Silicycle). The desired hydrochloride salt was precipitated from water/ethanol in approximately 80% yield and high purity (< 2% of the undesired impurities), even on a large scale (10).

Next, the hydrochloride salt of the fully deprotected penultimate intermediate was coupled with the desired side chain to prepare eravacycline. The reaction was carried out in acetonitrile and water. To achieve complete conversion, several charges of the acid chloride were necessary, and it was also found that adjustment of the pH from approximately 3 to approximately 7 after the second charge aided the complete dissolution of the starting material, allowing the reaction to go to completion. After the completion of the coupling, the pH of the reaction solution was brought to pH 6.8 to ensure hydrolysis of any over-acylated compounds to the desired tetracycline product.

Eravacycline was extracted using dichloromethane at pH 7.4. As an added benefit, the researchers found that the undesired C-4 epimer was partly removed in the aqueous layer and when the dichloromethane solution was dried with sodium sulfate prior to evaporation, thus increasing the purity of the tetracycline product (10). Finally, the bis-hydrochloride salt of eravacycline was prepared using an ethanol−methanol mixture containing an excess of hydrogen chloride and precipitated with addition of ethyl acetate.

“While some of the steps presented challenges, this overall route to eravacycline has enabled the production of sufficient quantities of the API for clinical testing. This tetracycline derivative has completed Phase II clinical studies and has been shown to be active against multidrug resistant bacteria and is therefore a candidate as a broad spectrum antibiotic for serious hospital infections. We are continuing to improve the process for future larger-scale manufacturing and are also developing an isolation procedure that will be suitable for commercial production of eravacycline,” Ronn notes.

References
1. D. J. C Constable et al., Green Chem. 9 (5) 411-420 (2007).
2. S. Woodward, Synlett. 10, 1490-1500 (2007).
3. A. Novak et al., Tetrahedron Lett. 47 (32) 5767-5769 (2006).
4. B. Kallolmay et al., Angew. Chem. Int. Ed. 44 (15) 2232-2234 (2005).
5. Y. Mata, J. Org. Chem. 71 ( 21) 8159-8165 (2006).
6. T. Cooper et al., Adv. Synth. Catal. 348 (6) 686-690 (2006).
7. A. Alexakis et al., Chem. Commun. 22 2843-2845 (2005).
8. D. Glynn et al., Tetrahedron Lett. 49 (39) 5687-5688 (2008).
9. M.G. Charest et al., Science 308 (5720) 395-398 (2005).
10. M. Ronn et al., Org. Process Res. Dev. 17 (5) 838-845 (2013).
11. J. D. Brubaker and A. G. Myers, Org. Lett. 9 (18) 3523-3525 (2007).