Strange Brews

There are very few things mankind has done which nature hasn't done first - and often better. The most advanced aircraft are primitive compared with insects and birds. The fastest computers are glacially slow compared with the human brain. And the most cunning catalysts, elegant engineering and smartest sensors can't make human-designed chemical processes anywhere near as efficient as the reactions carried out in nature.

Natural process produce the most complex molecules from simple starting materials, under mild conditions, with near-total selectivity and high yields. Human process engineers are increasingly adopting the attitude of `if you can't beat `em, join `em', and seeking ways to incorporate biological processes into production plants.

This is far from a new trick, of course; it's one of the oldest in the book. It's been argued that fermentation was one of the key technologies behind the development of civilisation. But the gathering pace of biotechnology, with its promise of engineered microbes acting as mini-chemical plants, is forcing the pace of development. As with other chemical processes, process engineers are facing the twin imperatives of maximising output while minimising both expenditure and waste; but this time, they also have to keep the precious micro-organisms alive.

It's the pharmaceutical industry which is leading the charge for new fermentation techniques. According to Peter Dunnill, research director at University College London's Advanced Centre for Biochemical Engineering, `the production, purification and formulation of new potential medicines poses major bioprocess challenges.'

The difficulties of processing plasmid genes - a human gene inserted into a ring of DNA to synthesise a particular protein, often as a vaccine - into an industrial plant are far from trivial. This, Dunnill explains, is because of the sensitivity of these - in biological terms - very large entities to mechanical forces.

The UCL centre specialises in studying the effect the `engineering environment' has on living cells and their macromolecular products. UCL is looking at accelerated methods of process design, verified in its pilot plant suite. This uses `scale-down' of key process steps to obtain accurate data for bioprocesses, which then helps companies minimise costly and lengthy pilot plant trials when combined with process modelling.

The Centre is also investigating the field of `metabolic engineering' to produce these biopharmaceuticals in fermenters. `This adapts the now-established techniques of genetic engineering, where a single human gene is incorporated into a bacterium to synthesis human therapeutic enzymes,' explains Dunnill. `In the newer approach, a complete set of genes is altered in such a way that the corresponding set of enzymes is also modified. This allows the synthesis of novel antibiotics, which can address drug-resistant infections.'

The genetics needed for these feats are complex, Dunnill concedes. However, they can be guided by engineering modelling, which is similar to that used in the other sections of the Centre.

Despite the exotic materials used, the problems to be overcome in fermentation are often familiar from other parts of the process industries. For example, the metabolism of bacteria which produce lactic acid slows down in the presence of lactic acid. Bob Lovitt of the University of Swansea's bioprocess technology centre believes that this problem could be handled by using membrane reactors. `These allow the bacteria to grow to very high cell densities, but they also maintain the flow of liquid through the reactor which carries away the lactic acid,' he explains. The high cell density means that the reactor is more efficient than conventional fed-batch reactors, but the constant flow and low level of lactic acid keeps the cells healthy and productive.

Membrane reactors can also be used for a variety of processes. `At the moment, we're only working with lactic acid, but this could also be used for yeast fermentation, which is inhibited by high concentrations of sugar,' he says.

Even more significant is its potential for fermentations involving E. coli - one of the key bacteria used in genetic modification. Normally, E. coli processes are carried out in a fed-batch system, where the bacteria's growth is controlled by oxygen concentration, levels of glucose and pH. The system is difficult to control and generally inefficient, says Lovitt. Membrane reactors could achieve the same results with higher cell density and, therefore, better conversion rates, he conjectures.

However, the reactor design is only half of the problem. `You need to know a lot about the kinetics of these systems, and I'm working with companies which are involved with lactic acid production to develop new technologies to manage the feeds and filtration.'

Developing processes to maximise yield is going to be vital for biotechnology, says Lovitt. `Intensive processes are going to be the way to go - we're talking about reactors of no more than a cubic metre in volume,' he says. These reactors produce little waste. `The waste from these reactions tends to have a high biological oxygen demand, so the less of it you have to treat, the better,' Lovitt comments.

Containment is an issue for all biological processes, but it's particularly pressing with fermentation involving genetically-modified bacteria. Everything must be sterile, and no GM material can be allowed to escape from the process. The techniques involved are not complicated, but the costs are high. Industry insiders believe this is likely to feed the trend of outsourcing - with pharmaceutical firms contracting production to fermentation specialists. As yet, this is a trend that has yet to take hold in the UK - a definite niche for companies with expertise in this area.