1: Introduction and Aims of the Book

Biocatalysis, by which we mean the use of enzymes and microorganisms as catalysts for chemical transformations, has had a long and distinguished history, particularly in the brewing, baking and animal feed industries. In addition, the development of fermentation-based processes for antibiotic production relies squarely on the fundamental properties of enzymes for the production of natural products. In the latter half of the 20th century, biocatalysis started to be viewed as a potential route to unnatural or synthetic compounds. For example, scientists in Schering in Berlin in Germany discovered that semi-synthetic steroids could be manipulated with exquisite selectivity using enzymes. In the 1970s, the use of hydrolytic enzymes (e.g. lipases and proteases) and oxidoreductases (e.g. alcohol dehydrogenases) started to gain popularity for the synthesis of chiral building blocks, particularly within the pharmaceutical industry where stereochemical purity in the final product was a high priority. The discovery that some enzymes (e.g. lipases and proteases) could be used under low water conditions in organic solvents extended the range of these biocatalysts to non-polar organic substrates. However, by the early 1990s, the range of biocatalysts being employed by organic chemists was still relatively narrow and confined to hydrolases and oxidoreductases. A typical international conference on biocatalysis at that time would focus ca. 80–90% of the presentations on these two classes of enzymes.

In the 1990s, several major developments occurred within the field of biocatalysis that together changed the face of the discipline and resulted in a rapid diversification of the enzymes available to synthetic chemists.¹  Firstly the use of molecular biology protocols as tools to clone, express and manipulate novel genes, and hence produce new enzymes, became much more widespread and meant that there was increasingly less reliance on the use of wild-type crude cell extracts for biocatalysis. Secondly, the use of protein engineering and directed evolution emerged as powerful strategies for optimisation of the properties of enzymes, particularly in the context of improving stability, stereoselectivity, substrate tolerance and catalytic activity. Thirdly, the availability of sequenced genomes meant that there was an explosion in the number of gene/protein sequences that could be mined and used as the basis for discovering new enzymes. At the beginning of 2000, the cost of DNA sequencing and gene synthesis began to drop considerably, as a result of technological advances, which meant that ordering synthetic genes became cost effective and a rapid way of generating novel biocatalysts for screening. Together, these new technologies resulted in a significantly broader range of enzymes becoming available to organic chemists. Nowadays, we have a plethora of different methods and approaches available to us for the discovery and development of new biocatalysts (Figure 1.1). The challenge has shifted from simply discovering new enzymes to trying to curate the existing databases in order to try to understand which sequences might code for which enzyme activity. De novo protein design has made major strides forward, and allows us to generate new enzyme activities from alternative scaffolds. Enzyme evolution has developed to the point where it is now routinely applied to optimise enzyme performance and represents one of the most powerful algorithms available to those interested in the development of new biocatalysts. Other scientific advances in the scale-up of biocatalytic processes provided greater levels of confidence that once a biocatalyst had been identified for a specific chemical transformation, there was a good chance of developing a practical large-scale process. Biocatalysis is now a maturing technology for the manufacture of a range of chemical products across a wide range of industries from pharmaceuticals and biofuels to polymers and personal healthcare products. Compared to traditional synthesis methodologies, biocatalysis offers a number of advantages that can contribute to more sustainable manufacturing processes. In addition, the emerging field of synthetic biology offers the potential for increasingly efficient synthesis by combining multiple biocatalytic reactions all within a single host organism.

Today, biocatalysts are increasingly being considered as an option when planning synthetic strategies for the construction of target organic molecules, particularly in cases where it is important to try to achieve some type of selectivity (e.g. stereo-, regio-, and chemoselectivity). The use of biocatalysts as an alternative to chemical reagents and catalysts can also provide benefits in terms of a reduced number of synthetic steps, lower costs of goods, reduced use of harmful solvents and improved safety profiles. Together, these benefits can lead to a more sustainable overall process with a lower overall environmental impact. However, a major barrier remains preventing the more widespread application of biocatalysis, namely a general lack of awareness and understanding amongst the synthetic organic community regarding the types of biocatalysts that are available and the various ways in which they can be applied in target molecule synthesis.

During the past ten years, a number of excellent books and reviews have been published highlighting the various biocatalysts that are now available and detailing the types of transformations that can be accomplished.²  Typically, these books are organised according to different classes of enzyme (e.g. hydrolases, dehydrogenases, transferases and lyases) and the various reactions are presented based upon the type of synthetic transformation that can be achieved (e.g. hydrolysis, reverse hydrolysis, C–C bond formation, C–X bond formation, oxidation, etc.). This approach provides the reader with an excellent overview of where specific biocatalysts can now be applied in organic synthesis, together with an insight into the tolerance of enzymes for unnatural substrates, which allows consequently for the synthesis of non-natural target molecules. However, as more new biocatalysts become available for application in synthesis, inevitably the question arises as to what is the best way of preparing a particular target molecule, or a building block with specific arrangements of functionality and chirality, given the fact that there may be several available options. This situation is particularly true for the synthesis of chiral alcohols, amines, carboxylic acids, esters, amides, etc., for which there are now several available biocatalytic systems that could in principle be used, and certainly many more than were available ten years ago. As more distinct biocatalytic platforms are developed, this expansion of options will continue and indeed accelerate. Indeed, our ability to discover new biocatalysts, and subsequently engineer them for broader substrate tolerance or new reaction chemistry, is rapidly increasing due to advances in technologies such metagenomics, genome sequencing, protein engineering, directed evolution and high-throughput screening.

Despite the now widespread application of retrosynthesis in everyday organic synthesis, there has been no systematic attempt to include biocatalytic reactions in the whole process of synthetic planning and design. This omission is more remarkable in view of the increasing interest in using biocatalysts for synthesis in both industry and academia. Students are typically taught retrosynthetic analysis at around the same time as more traditional C–C bond forming methods. As a result, the connection between biocatalysis and synthesis tends to be made towards the end of their studies, generating the perception that enzymes are specialised and not suitable for mainstream organic synthesis. If biocatalysis is to be fully incorporated into the synthetic chemist's toolbox, then the teaching of biocatalytic methodology and retrosynthesis must be concurrent, as it is with more traditional organic synthesis.

In the mid-1960s, synthetic organic chemists began to apply the principles of retrosynthetic analysis during the planning of new routes to target molecules. Retrosynthetic analysis, which was originally proposed by E. J. Corey, fundamentally changed the way in which synthetic chemists conceived and developed approaches to the total synthesis of natural and unnatural products.³  An important principle of retrosynthetic design is that each disconnection in the reverse sense must represent a feasible transformation in the forward synthesis direction. Students are generally taught to disconnect molecules by a formalised process of bond cleavage of either C–C or C–X bonds in the reverse direction. In addition, functional group interconversions (FGI's) are used to prepare the molecule for the desired bond cleavage. An important by-product of this method of analysis has been that synthetic chemists have discovered disconnections for which the corresponding forward synthesis methodologies were unknown and new additions to the synthesis toolbox have been developed as a result.

The invention of new reagents and increasingly catalysts for selective organic synthesis has resulted in the situation where today there is (almost) no molecule that is beyond the reach of the synthetic organic chemist. Organic synthesis using synthetic reagents and catalysts is the predominant technology for the manufacture of a broad range of chemical products, ranging from pharmaceuticals and agrochemicals to polymers and plastics, and including a diverse range of specialty chemicals in between (Figure 1.2). Organic synthesis by and large uses feedstocks derived from the petrochemical industry and indeed to a large extent the two disciplines have ‘co-evolved’ – one provides the starting materials for the other – and new chemical reagents and catalysts are designed to be compatible with oil-derived starting materials.

Figure 1.2 presents two alternative scenarios for undertaking organic synthesis. As noted earlier, many semi-synthetic pharmaceuticals are derived from natural products derived from fermentation processes. Over millions of years, Nature has evolved complex biosynthetic pathways for the production of a truly diverse set of natural products, including alkaloids, polyketides, terpenes, carbohydrates, amino acids, etc. Even some of the simpler natural products, such as the alkaloid (−)-sparteine, are fantastic examples of the way in which biosynthetic pathways are able to convert simple building blocks, in this case cadaverine, to structurally more complex target molecules using simple chemical transformations (Figure 1.3). Inspection of this pathway reveals a typical scenario in the biosynthesis of natural products, i.e. the enzymes involved are catalysing simple chemical transformations, e.g. oxidation, reduction, and hydrolysis. The bond forming and bond breaking processes involved are not sophisticated, but it is the way in which the reactions are orchestrated that results in the rapid construction of molecular complexity from simple achiral precursors. And of course, all of the reactions are subjected to catalysis by an enzyme – that is Nature's way. Imagine being asked to make a molecule such as (−)-sparteine in the laboratory using only catalysts for the synthesis! If you are truly to emulate Nature, then you would not be allowed to use protecting groups. Moreover, imagine being given an exam question in which you are asked to generate a synthesis of (−)-sparteine using cadaverine as the starting material!

However, despite the abundance of natural products and biosynthetic pathways in Nature, the enzymes involved in secondary metabolism have not found their way in a general sense into the organic chemist's toolbox. There are a number of reasons for this situation. Firstly, enzymes of secondary metabolism often have relatively low activity, probably as a consequence of the fact that natural products are often produced in low concentrations and hence there is no need for enzymes with high catalytic turnover. Secondly, these enzymes are often quite specific for particular intermediates along the biosynthetic pathway and hence do not lend themselves to being applied in a general sense to a broader range of substrates. Biosynthetic enzymes have, however, provided an enormously rich playground of molecules and associated enzymes in order to gain fundamental understanding of the enzyme mechanism and the ways in which complex molecules can be assembled.

This book is about exploring a ‘third’ approach to organic synthesis, namely the use of engineered biocatalysts. Unlike natural product biosynthesis, these enzymes will predominantly be derived from primary metabolism, and hence should have inherently higher catalytic rates since their natural function is to process large quantities of building blocks and intermediates that are essential to cellular metabolism (e.g. amino acids, carbohydrates, and nucleotides). The challenge here is to develop an increasingly broader range of distinct biocatalysts such that a wider range of organic molecules come into play and become potential targets for biocatalytic synthesis. Clearly at the same time as the biocatalytic toolbox expands, it is important to gain a greater understanding of when and how to apply these new engineered enzymes, hence the need for ‘biocatalytic retrosynthesis’.

The biocatalytic toolbox has expanded significantly during the past ten years to the point where today more than 50 different classes of enzymes are commercially available, offering chemists an attractive alternative to chemical reagents or chemocatalysts for chemical transformations. Moreover, given the current rate of development of new engineered biocatalysts, it is likely that this number will double in the next few years. At present, the use of biocatalysts is focused primarily on functional group interconversions (FGI's) for the preparation of (chiral) building blocks and intermediates. Thus, there are now myriads of applications, including in industry, of the use of ketoreductases (ketone to alcohol), lipases and esterases (esters to acids and the reverse), and increasingly transaminases (ketone to amine). However, other areas are less well developed, including C–C and C–X (X=N or O) bond formation, as well as selective oxidation reactions for remote C–H activation leading to hydroxylation and other types of functionalisation. The current focus on a relatively small sub-set of chemical transformations is partly due to the lack of available biocatalysts, although this situation is rapidly changing, but also due to a lack of awareness.

Based on the reasons outlined above, we believe that it is now timely to consider the development of guidelines or rules for “biocatalytic retrosynthesis”,⁴  in which molecules are disconnected with consideration of applying biocatalysts, as well as chemical reagents & chemocatalysts, in the forward synthesis direction. We believe that such a development will stimulate both new ways of teaching the art and science of biocatalysis and also research into areas for which new methodology is required. In its most elementary form, biocatalytic retrosynthesis might simply involve replacing a chemical reagent used for a particular transformation with a biocatalyst. This type of scenario is illustrated in Figure 1.4 for the manufacture of the drug montelukast. Montelukast is now a generic active pharmaceutical ingredient (API) and has been the subject of several attempts to reduce the cost of its production. One of the original manufacturing routes to montelukast employed the reagent (S)-(DIP)-chloride to carry out the key step of asymmetric reduction of the ketone to give the enantiomerically enriched chiral secondary alcohol. Recently, the use of ketoreductases with cofactor recycling has become a more favoured route and indeed variations of this process are now operated in several countries for montelukast production.⁵  The table in Figure 1.4 provides a comparison of the chemical and biocatalytic processes with respect to various different parameters, such as substrate loading, enantioselectivity, solvent usage and waste. This comparison highlights both the significant differences that can be encountered when comparing chemical and biocatalytic processes, together with the reasons why the biocatalytic process may be preferred in terms of improved solvent regime and a reduction in overall waste.

However the full impact of biocatalytic retrosynthesis can only be realised when the introduction of one of more biocatalytic steps into a synthesis enables a major redesign of the synthesis of the target molecule. The use of a biocatalyst might result in a shorter route, with a different and less expensive starting material, and might also lead to greater selectivity and reduced waste. The aim of this book is to explore such ideas and develop a better understanding of how the use of biocatalysis might effect such a step change in target molecule synthesis.

An increasing opportunity for biocatalysis is the development of cascade processes in which two or more engineered enzymes are combined in vitro or in vivo to enable multi-step conversion of simple starting materials to more complex targets.⁶  Figure 1.5 illustrates how biocatalytic retrosynthesis can be integrated with cascade reactions to assist with the planning and development of multi-enzyme transformations. In this approach, the target compound D is subjected to a detailed analysis to identify the required biocatalysts, which are then engineered individually followed by construction of the cascade, either in vitro or in vivo. This cascade is then subjected to rounds of optimisation in order to achieve the target yield of the product. Such design–build–test systems will increasingly feature in the future, particular as the full power of synthetic biology is leveraged.

In summary, the aim of Biocatalysis in Organic Synthesis: The Retrosynthesis Approach is to (i) illustrate the current applications of biocatalysis using worked examples and case studies and (ii) develop new guidelines for identifying where biocatalysts can be applied in organic synthesis. The book contains a complete description of the current biocatalyst classes that are available for use and also suggests areas where new enzymes are likely to be developed in the next few years. The book also contains worked examples that enable the reader to practice disconnecting target molecules to find the ‘hidden’ biocatalytic reactions that can be potentially applied in the synthetic direction.

Chapter 1 outlines the aims of the book and briefly reviews the current landscape that has resulted in a significant increase in the application of biocatalysis, particularly for the synthesis of pharmaceuticals, fine chemicals, agrochemicals, polymers, flavours & fragrances (including for nutraceuticals and personal healthcare). In this chapter, the concept of ‘biocatalytic retrosynthesis’ is introduced as a means of identifying opportunities for introducing enzymes into syntheses in order to redesign synthetic routes to target molecules.

Chapter 2 provides a basic introduction to enzymes, including their structure, production, properties and considerations for their use, and their potential advantages over chemocatalysts. This chapter also briefly covers biocatalytic technologies and some practical aspects, e.g. isolated enzymes, immobilised enzymes, whole cells, cofactor recycling, protein engineering and directed evolution of enzymes, and high-throughput screening.

Chapters 3–9 provide a detailed account of the principal classes of different biocatalysts and how they can be used to synthesise particular motifs or building blocks that are of interest in target molecule synthesis. Information can also be found here regarding the mechanism of particular enzymes, especially as it relates to their application.

• Chapter 3 – Hydrolysis of esters (lipases, esterases), amides (acylases, proteases), nitriles (nitrilases), epoxides (epoxide hydrolases), sulfate esters (sulfatases) and carbon–halogen bonds (dehalogenases).

• Chapter 4 – Reverse hydrolysis, including the application of hydrolytic enzymes for the synthesis of esters, amides, lactones, lactams, etc.

• Chapter 5 – Reduction of CO (KREDs, ADHs), CC (ene reductases), CN (imine reductases, amine dehydrogenases, amino acid dehydrogenases), CO₂H (carboxylic acid reductases), nitroaromatics, azides, NN, sulfoxides and N-oxides.

• Chapter 6 – Oxidation of C–H bonds (P450 monooxygenases, peroxygenases), Baeyer–Villiger monooxygenases, epoxidation (peroxidases, monooxygenases), heteroatom oxidation, allylic oxidation (lipoxygenases), oxidation of amines and alcohols using amines/alcohol/amino acid oxidases, C–Hal bond synthesis (halogenases, fluorinases), oxidation of imines and aldehydes (xanthine DHs and aldehyde oxidases) and dealkylation of N-, O-, and S-ethers.

• Chapter 7 – C–X bond formation, including C–N bond synthesis (transaminases, ammonia lyases) and C–O bond synthesis (fumarase, hydratase).

• Chapter 8 – C–C bond formation including aldolases, TPP-dependent lyases, cyanohydrin lyases, alkyltransferases, Pictet-Spenglerases, carboxylation, cyclopropanation (engineered P450s) and terpene cyclases.

• Chapter 9 – Miscellaneous biocatalysts, including racemases (amino acid, mandelate).

Chapter 10 introduces the idea of developing a structured approach to the disconnection of target molecules based upon biocatalytic retrosynthesis. The reader will be guided through the various disconnections that are possible, both for acyclic and cyclic systems, in order to gain an understanding of where biocatalysts can be applied in organic synthesis. The various disconnections possible are organised into one of five different groups:

• Chapter 10.1 – Acyclic systems: substituted alcohols, amines, carboxylic acids, ketones, etc. (1 functional group).

• Chapter 10.2 – Acyclic systems: 1,2-, 1,3- and 1,4-diols, hydroxycarbonyls, dicarbonyls, etc. (2 functional groups).

• Chapter 10.3 – 4-, 5-, 6-, and 7-membered carbocyclic rings.

• Chapter 10.4 – 4-, 5-, 6-, and 7-membered rings containing one or more heteroatoms.

• Chapter 10.5 – Substituted aromatic and heteroaromatic rings.

Chapter 11 provides a comparison of different biocatalytic routes to target molecules. The reader is taken through a series of well-known pharmaceutical targets where multiple different biocatalytic routes have been developed and in some cases scaled up for commercial applications. The aim of this chapter is to begin to gain some insight into the way in which target molecules can be disconnected back to simpler precursors, which can then be transformed using biocatalysis.

Chapter 12 concludes the book by giving readers the opportunity to test their understanding of biocatalysis and gain experience in disconnecting target molecules based on the principles of biocatalytic retrosynthesis. 25 worked examples (with answers), of increasing difficulty, are provided to enable students to develop their skills and apply their knowledge.