Natural progression

Scientists were exploiting microbes and enzymes to make chemicals long before biotechnology achieved its current status. Dr Sarah Houlton reports on the latest contributions nature is making to the progress of the pharmaceutical industry

Biotechnology is perceived as a thoroughly modern science. It has come to prominence along with recent developments in biology like the unravelling of the human genome, and advances in genetically modified plants. But the power of biotechnology has in fact been exploited in chemical transformations for thousands of years — ever since yeast was first used in the brewing of beer.

Using organisms, cells and enzymes to make chemicals is nothing new either. Fleming discovered penicillin in the mould Penicillium rubrum back in 1928, and fermentation remains the standard method for the production of β-lactam subunits for the vast range of semi-synthetic penicillin and cephalosporin antibiotics. And nowadays the power of nature is increasingly being harnessed to produce pharmaceutical actives and intermediates.

Because nature is chiral, biotechnology-based processes have a big advantage: they invariably give chiral products. But by and large companies used to be afraid of using enzymes and microbes routinely in chemical synthesis because they felt the perceived difficulties far outweighed the potential benefits.

However, the recent rapid growth in the importance of single enantiomer drugs means that the ability to introduce chirality into molecules reliably and in high yield is now an essential skill for chemical companies supplying pharmaceutical intermediates and ingredients. And the number of contract manufacturers advertising that their capabilities include biotechnological reactions has grown rapidly over the past two or three years.

catalysts

Enzymes are finding increasing use as biological catalysts because of their stereospecificity. A good biocatalytic process can often replace three, four, or even more chemical steps, as its high substrate specificity can often remove the need for protection and deprotection procedures. The chemistry is carried out under mild conditions, and elevated temperatures and noxious chemicals are unnecessary.

They are not cheap, but if enzymes are used in the synthesis of high value products, if the reaction is much more efficient, or if they can be reused, the economic benefits become clear. Recent developments in enzyme chemistry have made their use on an industrial scale more straightforward, and various companies are working on ways of designing specific activities into enzymes.

Enzymes are one of the types of protein that genes create to perform specific functions. The DNA in genes is the blueprint for protein synthesis: it first forms RNA by transcription, and this is the template for protein synthesis by translation. Each sequence of three nucleotide bases, or codon, is the code for one of the 20 amino acids, and the residues are strung together following the order stipulated in the RNA, until a further codon telling the synthesis to stop is reached.

isolation less than simple

As reagents enzymes are not easy to handle, and isolating products from the reaction soup can be less than simple. But if they are immobilised in some way, then the isolation immediately becomes much less difficult, as the enzyme can be filtered off and, often, reused.

An example is the enzyme penicillin G amidase, which is used in the production of 6-aminopenicillanic acid, the core β-lactam subunit of penicillins.

A similar reagent, cephalosporin C acylase, is used to make 7-amino cephalosporinic acid, the moiety at the heart of semi-synthetic cephalosporins. As the only real alternative would be a full chemical synthesis, using these enzymes to modify the penicillin G or cephalosporin C created by fermentation is much more cost effective, as it uses nature to introduce the complex functionality as well as the chirality.

Many enzymes are now commercially available for use in chemical processes. But the best enzyme for a specific process is not necessarily available off the shelf, and a number of speciality enzyme companies offer a range of methods for customising enzymes. As enzymes originate from DNA, then modifying the DNA in genes will lead to modified enzymes. Several processes are used for this.

US-based MaxyGen, of Redwood City, California, uses its proprietary gene shuffling technology to create novel enzymes. An enzyme with some activity is pinpointed, and the DNA that created it is modified by swapping some of the genes around. The new enzymes are then screened for activity, and, as the company's Tassos Gianakakos explained at this year's CPhI conference in London, it typically takes two or three rounds of shuffling before truly useful activity is obtained. 'You could go as far as eight or nine rounds of shuffling,' he said, 'and you see exponential increases in activities in the early stages.'

The company is focusing on the generation of meaningful diversity — creating enzymes that perform relevant chemistry, without being limited by natural diversity. 'We prefer to start with regular run-of-the mill-genes,' Gianakakos said, 'and shuffle them to give the desired characteristics, rather than spend ages looking for a successful natural enzyme.' He added that the company has applied the technology to around 40 projects so far, and has been successful in all of them. 'It is a rapid, cost-effective way of generating new catalysts custom-designed to fit the process,' he claimed.

A different approach is taken by US enzyme specialist Diversa, based in San Diego, California. It has a huge library of DNA derived from organisms. Its thousands of libraries each consist of up to a billion fragments of DNA, and these pieces are screened for activity in a particular reaction.

The numbers involved are so huge that high throughput screening techniques are essential. 'We can screen up to a billion genes a day,' explained senior director, business development Dr Patrick McCroskey. Once a hit is found, the active DNA fragment is identified, sequenced, cloned and extracted to give an enzyme tailor-made to catalyse a specific reaction.

Like MaxyGen, Diversa has proprietary technologies for altering the activity of the enzymes. It has several molecular evolution technologies to change the amino acid sequence in the enzyme, whether a single site or random mutagenesis changes for more wholesale alterations.

A further presentation at the CPhI conference gave an insight into another company's enzyme strategy. Thomas Daußmann from Juelich Fine Chemicals, a spin-off company from the University of Juelich in Germany, explained how his company has discovered several enzymes from natural sources with potential industrial applications. For example, an enzyme isolated from Lactobacillus brevis has been found to have very broad substrate specificity, as well as having a relatively high thermal stability. βb-Diketones and b-ketoesters are both commonly found in the structures of drug molecules, and Juelich has found its L. brevis-derived enzyme can be used to synthesise chiral b-ketoesters from b-diketoesters. A stereoselective two-step chemoenzymatic reduction sequence gives access to all four diastereomers of the doubly reduced diketoester. (Scheme 1). In the first reduction, baker's yeast gives one enantiomer, and the alcohol dehydrogenase isolated from L. brevis the other. These can then be further reduced using syn and anti selective borohydride reductions to give all four possible diastereomers, all of which could have potential as pharmaceutical intermediates.

One problem with enzyme-catalysed reactions is that they need some form of cofactor to regenerate the enzyme within the reaction mixture. In nature, this cofactor is generally something like NADH, but to use this stoichiometrically at process scale would be prohibitively expensive. Degussa has addressed this problem by developing a method for regenerating the cofactor in situ.

tonne scale production

As the company's Dr Harald Groeger explained, a second enzyme, formate dehydrogenase, can be added to the mixture along with ammonium formate, which is digested as NADH is regenerated from NAD+ (the second enzyme's cofactor). This means that the reaction shown in Scheme 2, the transformation of an a-keto acid to an L-amino acid can be carried out enzymatically, using two different enzymes in the same pot. This chemistry is now carried out at tonne scale.

Using isolated enzymes as catalysts is not the only way of exploiting the power of biochemistry in synthesis; whole organisms can also be employed. The reaction mixtures are more complex than those using an isolated, immobilised enzyme but, sometimes, the chemistry a microbe can perform cannot be done by an isolated enzyme.

Avecia Pharmaceuticals' Bob Holt spoke about some of the work his company has been doing in the area of microbial synthesis of chiral intermediates, particularly the resolution of racemates. A strategy for resolving amines was developed, which involved making the amide and hydrolysing it. Various commercially available enzymatic systems were screened, but all those tried gave slow reactions and poor enantioselectivity, so the company looked for a novel microbial system that would give better results.

An Arthrobacter species that generates an amidase was identified. This can cleave the amide bond in one enantiomer, but not in the other, thus allowing the separation of the racemate into its two enantiomers. An example, shown in Scheme 3, is the resolution of a potential intermediate for the antiviral drug indinavir (Crixivan).

Avecia has also been using a microbial DMSO reductase to catalyse sulphoxide resolutions. The enzyme is involved in the final step of anaerobic respiration in enteric bacteria. The company discovered that some of these bacteria would accept racemic sulphoxides, notably E. coli and P. vulgaris, which selectively gave the S enantiomer, and R. capsulatus which selectively gave the R enantiomer.

The importance of biotechnological processes in organic synthesis can only increase. Not only can they provide an efficient route to chirally pure intermediates for further elaboration into active drug substances, but with the increasing numbers of biopharmaceutical products being developed as a result of the information gained from the human genome project, they will surely become essential. Complex biological molecules are invariably expensive and long-winded to make by purely chemical means, and enzymes and microbes will undoubtedly become more important in the synthesis of these products.

  Using biotechnology in synthesis is a likely addition to the toolbox of available technology, with the best fit for each particular step of a synthesis being chosen, whether it is a chemical or a biotechnological process.

As time is money for a pharmaceutical company with a potential blockbuster drug on its hands, and only a limited patent life associated with it, the temptation is to stick with chemical reactions to save time and money when compared with a biotechnological process.

However, as pharma companies and contract manufacturers gain experience in what sort of chemistry enzymes and organisms are good at, it should be easier to spot early on the stages of a synthesis that might be more efficient with a biotechnological reaction.

Designing appropriate biotechnology in at the beginning of the development process could be the key to optimising the efficiency of an industrial synthesis.

Metanomics investigates genetic potential of plants

Nature provides not only the ingredients for chemical reactions. Many useful ingredients are isolated directly from plants, from highly potent actives like the anticancer drug paclitaxel (Taxol), found in tiny amounts in the bark of Pacific yew tree Taxus brevifolia, to bulk excipients such as starches from sources including maize and potatoes. Genetic engineering is being used to harness the power of growing plants and increase the yield of important constituents for isolation. Genome sequencing is an important step on the road to creating functional plants, but it is only the beginning. It merely provides the code that is the blueprint of life, but without decoding it to establish precisely what each gene does, it is of little practical use. This is where functional genomics comes in. One company that is putting huge resources into working out the functions of genes is Metanomics. The company was founded in 1998 as a joint venture between BASF and the Max Planck Institute in Golm, near Potsdam in Germany. Plants have an estimated 25,000 genes, compared with 1,500 for bacteria, 6,000 in yeasts and 34,000 in humans. Yet the precise functions of only around 10% of these genes are known. Being able to exploit the genetic potential of plants to the full will mean elucidating the precise functions of the remaining 90%. Metanomics is looking at the action of individual genes directly on the metabolism and performance of Arabidopsis thaliana plants. The company is using high throughput screening processes, analogous to those used in modern drug discovery labs, to investigate the complex interactions between genes, their functions and the regulation of these functions. Its work is based around the field weed mouse ear cress, Arabidopsis thaliana, which has several advantages: its genetic blueprint has been established; it contains almost all of the genes present in crop plants; and it grows rapidly. Metanomics overexpresses (i.e. adds) genes from other organisms to see what happens, giving important information about the function of the different genes. As the company's managing director, Dr Arno Krotzky, explained at a biotechnology conference held by BASF in Berlin in October, the company is collecting and evaluating the analytical signals occurring during a metabolic analysis. Targeted methods are used to investigate changes in plant constituent showing special commercial or scientific interest, eg, vitamins, oils, sugars and amino acids," he explained. As well as genes that could be used to improve the agricultural properties of plants, such as tolerance to drought, the company has discovered genes that fundamentally influence the content of important plant constituents such as healthy fats and oils, and initial patent applications have been filed for these genes. BASF has already engineered a plant that produces high levels of vitamin E but, ultimately, one can imagine plants being designed that could make complex naturally derived actives like paclitaxel in high yields. There is a long way to go before all functions and gene interactions are fully explained, but work like this could prove the key to producing important pharmaceutical ingredients by plants as they grow.