They can take the heat - are they in the kitchen?

Pasteurisation is one of many lasting contributions to human welfare made by the great French chemist Louis Pasteur, the anniversary of whose death in 1895 we celebrated recently. But hot on the heels of those celebrations comes news that pasteurisation is not as effective as was once believed. Research at the University of Warwick suggests that some bacteria thought to be killed by the process - in particular Escherichia coli, certain strains of which cause gastroenteritis - can recover afterwards and begin to grow again. Thus pasteurised milk and other foods may be less safe than we had imagined.

Pasteurisation differs from sterilisation. The aim is not to destroy all microbes, but simply to eliminate those responsible for infections such as tuberculosis, brucellosis and gastroenteritis, and those that cause rapid spoilage. In contrast to ultra-high temperature (UHT) treatment (which does sterilise milk, which can then be kept without refrigeration), pasteurisation does not impair taste. It is achieved by raising the temperature of milk to about 65C (much lower than that needed for complete sterilisation) for half an hour.

In recent years there have been several experiments suggesting that heat treatment designed to destroy pathogenic (disease-causing) bacteria in food may actually leave some cells alive, though in a condition in which this is difficult to prove. Rather than being killed, they suffer a type of damage known as heat stress. As a result, the organisms do not begin to grow immediately when a sample of the food is placed on nutrient jelly - the normal way of screening for the presence of bacteria. The basis of heat stress is not fully understood. However, it may reflect damage to the membrane surrounding bacterial cells, impairing the ability of the membrane to transport nutrients from outside into the cell.

Ken Flint and C-H Lim, at Warwick University, hoped to throw light on heat stress by holding bacteria at high temperatures for much longer periods than other investigators have used, but also to allow a lengthy period afterwards to see whether any of the cells recovered. They grew E coli on nutrient broth, washed the cells and suspended them in flasks of lake water that had been filtered and sterilised and which contained insufficient nutrients to support bacterial growth. The researchers then left the flasks at the pasteurising temperature of 65C for six hours (far longer than in normal pasteurisation). From time to time they removed small samples, which they placed on nutrient jelly to see how many bacteria grew.

In the final part of the experiment, Flint and Lim cooled down the flasks to 15C, then left them at this temperature for seven days. Again, they took small samples each day to find out how many, if any, bacteria appeared when the samples were added to nutrient medium.

The results from the first stage were predictable. The number of bacteria that grew on the nutrient jelly declined rapidly, so that no E coli could be detected even in relatively large samples after two hours. However, once the bacteria had been transferred to 15C they began to regain the capacity to grow on nutrient medium. Within two days of the flasks being cooled down, live bacteria could again be recovered from samples placed on the medium. After seven days at 15C, the flasks contained the same population density of viable bacteria as they had at the outset of the experiment.

"Given that pasteurisation is usually carried out at temperatures around 65C for only 30 minutes in the dairy industry, the recovery of E coli, a potential pathogen, could pose a problem for the food and dairy industry," Lim and Flint write in Letters in Applied Microbiology. "The ability of bacteria to recover from the effects of heat treatment has to be taken seriously."

This does not mean, of course, that we should look with suspicion or anxiety on the daily pinta. Had there been a major public health problem attributable to the deficiencies of pasteurisation, this would have come to light long before now.

Nevertheless, the warning from Warwick is timely - not least because it forms part of a wider mosaic of discoveries concerning our knowledge and ignorance of the hazards posed by disease-causing microbes. Over the past 10 years, scientists have gradually realised that there are many more microbes, and many more species of microbes, in the environment than is evident from the results of conventional tests to reveal their presence.

Living cholera bacteria, for example, can exist in a form in which they fail to grow on nutrient jelly in the laboratory. Likewise, the use of new techniques that amplify microbial DNA and probes that "recognise" its constituent genes have revealed hitherto unknown bacteria in river water that also cannot be cultured in laboratory media.

The health implications of such findings, like those of Flint and Lim, are as yet unclear. However, they now need to be assessed and to be considered alongside other changes in our methods of preserving foods. Thus the recent rise in food poisoning caused by listeria is partially attributable to the growth of commercial and domestic refrigeration - using temperatures at which listeria can grow more vigorously than many other bacteria. Bacteria are great opportunists, renowned for their capacity to exploit openings of this sort. Now we know that some of them may also be able to evade conventional methods of detection. It is a worrying combination.