Biotransformation by whole cell systems

Chiral Technologies outlines the benefits of whole cell biotransformations in chiral production using a new coenzyme developed for making chiral hydroxy ester.

There are many different ways of imparting chirality to a molecule. The technologies can be categorised under three headings: Resolution; Chiral Pool and Asymmetry. It is this latter technique of taking a pro-chiral molecule and transforming it quantitatively into a single chiral molecule that is of most interest to manufacturing chemists, and there are several asymmetric technologies available, with most interest centred on biocatalysis and metal catalysis.Biocatalysis is very stereo-specific and is widely used in the manufacture of chiral molecules for pharmaceutical intermediates. The technology has low variable costs and has the potential to make chiral molecules on a large scale efficiently and cost-effectively.

Biocatalysis is usually undertaken in aqueous systems at near neutral pH and at low temperatures and pressures. Several studies have confirmed that biocatalysis has the potential to offer a more environmentally friendly solution than metal catalysis or conventional chemistry by reducing solvents, heating and waste, and has a reduced effect on the planet when assessed by life cycle analysis.

Biocatalysis utilises reaction mechanisms that are used in natural plant and animal systems, such as:

1. Oxidoreductases - oxidation/reduction reactions

2. Transferases - transfer of functional groups

3. Hydrolases - hydrolysis reactions

4. Lyases - elimination of a functional group

5. Isomerases - isomerisation

6. Ligases - bond formation coupled with ATP hydrolysis.

Daicel has been developing biocatalysis and applying the technology to industrial biotransformations for more than 20 years. It has concentrated on the development of whole cell biocatalysts for the manufacture of molecules, such as chiral alcohols, chiral amines and chiral amino acids. A range of reactions is shown in figure 2. These reactions are generally asymmetric reductions and transferases, which will benefit from the advantages of whole cell systems and coenzyme regeneration.

stereoselectivity

The active species in biocatalysis is the enzyme. Enzymes are almost exclusively proteins and function by catalysing chemical reactions. In essence, the protein forms a three- dimensional structure that has specific functional grooves that can stereoselectively bind chemical species. The interaction with the protein assists in stabilising transition species and thus catalyses the reaction. Catalysis rates of up to 1015 are possible.

Within the area of biocatalysis there are several options available that may lead to some confusion. Biocatalysis is a relatively young technology and there are many emerging companies trying to make a presence by providing particular technologies for different stages of pharmaceutical development. At this time there are three possible ways of using biocatalysis:

• Isolated enzyme

• Immobilised enzyme

• Whole cell

The manufacturing chemist has to sift through the many promotional claims in order to make a selection from these technologies. The isolated enzyme is easily applicable in development laboratories and the desire to scale-up to manufacture may be seen as a logical progression. This is not always the best choice, however, as biocatalysis manufacturing has many variables that will affect the choice of biocatalysis medium. Some variables are shown in table 1.

Whole cell catalysis offers significant advantages, especially when used in reactions where a coenzyme is required. In all reaction areas, whole cell biocatalysis requires professional development of the enzyme; selection and growth in a suitable host cell and access to appropriate manufacturing equipment (scale and isolation).

enzyme selection

Japan has an extensive history of working in the area of biocatalysis. One reason for this has been attributed to the wide diversity of climates within Japan, which has given rise to a vast number of bio-organisms within a small geographic area.

Enzymes are found naturally in bacteria, fungus and yeasts. Such wild organisms may be able to undertake biocatalysis but will generally require some bioengineering to improve productivity, stability and selectivity. Each wild organism will contain several thousand enzymes and the isolation, selection and optimisation of an appropriate enzyme is a lengthy process. However, once an enzyme has been isolated and cultivated it can be applied to several other reactions. By this route an enzyme library can be prepared.

Daicel has an enzyme library comprised of enzymes that have been isolated, cloned, engineered, and re-introduced into a common host. The enzyme library consists of 100 enzymes expressed in a whole cell. The library can be further defined as 70 oxido-reductases (50 suitable for chiral alcohol synthesis) and around 30 enzymes capable of other reactions, e.g. trans-aminases. The time taken to screen a pro-chiral molecule against these library enzymes is about two weeks. In addition to this cloned enzyme library, Daicel has a library of 3,000 micro-organisms that can be used to screen chemical reactions for further reactions and selectivity.

Such a cloned enzyme library offers several advantages:

1. It is highly active; regio- and stereo-selective

2. It offers fast enzyme screening, because substrate specificity and stereo-selectivity for the library enzymes are well known

3. It offers easy scale-up because their host is common E. coli and culture, induction, bioconversion, and purification methods are standard

4. It is safe because the host E. coli K-12 strain is non-pathogenic to humans and animals and not viable in the environment.

For a worked example of bioengineering we will discuss development of a coenzyme for use in the manufacture of chiral hydroxy ester, as these are useful as intermediates in the synthesis of several pharmaceuticals. The stereo-selective reduction of a defined carbonyl group to a hydroxy in a dicarbonyl compound such as in ethyl (S)-4-chloro-3-hydroxybutanoate is a cost effective route to this intermediate used in the manufacture of intermediates for HMG-CoA reductase inhibitors.

The reduction of a carbonyl containing compound to the chiral alcohol requires donation of a proton to the system. The proton donor is usually a coenzyme (normally nicotinamide adenine dinucleotide, (NADH) or nicotinamide adenine dinucleotide phosphate (NADPH), which releases the proton to the system and becomes NAD or NADP). This coenzyme is very expensive. The use of a whole cell system offers the advantage of enabling the cell to be engineered to contain the enzyme that undertakes the reduction, as well as a second useful enzyme that will regenerate the coenzyme to its protonated state by abstracting a proton from a further small labile molecule, such as formic acid or glucose. A reaction scheme is shown in figure 3.

coenzyme regeneration

Asymmetric reduction by biocatalysis requires an efficient regeneration system of the coenzyme, NAD(P)H, especially, in recombinant E. coli cells. E. coli cells can be used in this reaction as they have poor coenzyme-regeneration activity.

The most frequently used coenzyme-regeneration method uses glucose dehydrogenase (GDH) because it can regenerate not only NADPH but also NADH. GDH has high activity in a broad pH range, with good stability to organic compounds. However, it has a major deficiency: gluconate is produced in an equimolar amount to that of the desired chiral compound, so the aqueous waste stream will contain gluconate.

NAD+-dependent formate dehydrogenases (FDH) have also been used for the regeneration of NADH The co-enzyme-regeneration method using FDH has several advantages:

1. Its reaction is irreversible

2. Formic acid as the substrate is a small molecule and inexpensive

3. The by-product is carbon dioxide which is easily removed

4. Its environmental burden is relatively low.

FDH, however, has generally not been applied to industrial manufacturing, as it has a low activity and is labile to heat, some metals, oxidation by air, and hydrophobic compounds, especially alpha halo-ketones.

new coenzyme

We decided to investigate the potential to bioengineer FDH so that it retained its environmental benefits but gained the stability and productivity, so far only seen with GDH systems.

We have compared FDH with GDH as an NADH-regenerator to produce a hydroxy ester with E. coli cells. The use of GDH yielded 45.6g/l of product in >99% e.e from 50g/l of starting material. However, the use of FDH yielded only 19.0g/l of the desired product in >99% e.e. The low productivity of FDH was assumed to be as a result of the low activity and instability of FDH against the starting material.

Our starting assumption for FDH improvement was that the modification of some of the 'Cys' residues would improve the stability of the enzyme to the starting material. We characterised the enzyme and subsequently examined the substitution of 5 Cys-residues (6, 146, 249, 256, and 355-Cys) by site-directed mutagenesis. This involved substituting alternative amino acids for the Cys. Amino acids that were substituted included Ala, Val and Ser. These mutant FDHs were evaluated by enzyme activity and the amounts of desired product produced.

Our conclusion was that the mutant, designated as McFDH-26, was the optimum enzyme as it provided the highest productivity whilst retaining excellent % e.e. McFDH-26 was evaluated for the productivity of transforming the keto ester on a mini-jar (one litre) scale and the amount of product attained was 49.1g/l in >99% e.e. McFDH-26 has substituted Cys 6 with Ala; Cys 146 with Ser and Cys 256 with Val. These results demonstrate that McFDH-26 has comparable manufacturing capacity to GDH when used as an NADH-regenerator.

A cloned enzyme library generally facilitates easy scale-up, because we can use a common host, usually E. coli, a common culture, induction and harvest method of cultured cell, as well as common or patterned reaction and purification method.

The manufacturing process for biotransformation requires specialised equipment, shown in figure 1. The process requires the batch or continuous growing of the selected micro-organism. This means taking a small sample of the micro-organism (that is expressing the desired enzyme) and duplicating the cell in a broth. The manufacture of the catalyst requires the addition of feed to the broth, such as nitrogen sources and sugars. When a suitable concentration and activity is achieved the cells are isolated (harvested) using techniques such as filtration and centrifugation. The quantity of cells produced is dependant upon the reaction scale, but generally a single broth will supply sufficient material for several biotransformations and is not rate limiting for the process speed.

larger vessels

The reaction vessel is significantly larger than the vessels used for chemical reaction. Whereas a chemical reaction will require a 0.5-3m3 reaction vessel, biotransformation will use a 1-5m3 reaction vessel due to the low concentration of cells that the liquor will support. Use of the larger vessels ensures that productivity is no lower than that expected from a chemical reaction.

At the end of the reaction, the reaction liquors require processing to remove all traces of cell proteins etc., and to provide the desired compound. This requires access to several pieces of filtration and isolation equipment.

The end process is robust, highly stereo-selective, scalable, cost effective and environmentally-friendly.