Friend or foe?
By Debora MacKenzie GIVE or take a few bits of mobile DNA, the bacteria Bacillus thuringiensis (Bt), B. cereus and B. anthracis are all the same species, biologists believe. If you’re thinking “so what?”, then consider this: Bt is sprayed over crops in vast quantities and B. anthracis is the bug that causes anthrax. Small genetic differences have so far maintained the distinction that makes B. anthracis a notorious human pathogen and Bt merely a useful pest control bug. However, Bacillus expert Lars Andrup of the National Institute of Occupational Health in Copenhagen has identified a novel gene-swapping system that enables Bt to exchange an unusually wide variety of DNA with other Bacillus cells. The potential for spawning very dangerous strains and unleashing them into the environment is clearly there, he says. So why use Bt at all? For one thing, it is a highly successfully pesticide. Bt makes a toxin that kills insects but hurts nothing else. Genes for the toxin have been engineered into crops, but most farmers, timber growers and gardeners get it from live bacteria. More than 500 tonnes—five billion billion bacteria—are sprayed annually in the US alone. Similar amounts are sprayed in Europe. It is the only designated insecticide permitted on organic produce in Britain. If outbreaks of anthrax had been traceable to cabbage patches we would have known about them. But before you consider the anthrax link, some claim that Bt even in its familiar form may not be as benign as we like to think. A closer look at its genes shows it is remarkably similar to B. cereus, an organism which causes about four serious outbreaks of food poisoning a year in the US. The only difference between them is a few plasmids. And it now appears that Bt is well equipped for swapping these small DNA loops with other bacilli. The surprising but generally held view of Bacillus specialists who met in New Mexico in August is that, plasmids aside, Bt, B. cereus and B. anthracis are one species. Take away its insect-killing plasmid, for example, and “Bt cannot be distinguished from B. cereus,” says microbiologist Anne-Brit Kolst of the University of Oslo. “All three are one species based on genetic evidence.” Paul Jackson of Los Alamos National Laboratory, New Mexico, says some strains of B. cereus are more different from each other genetically than they are from B. anthracis. Bt, B. cereus and B. anthracis were considered different species because they favour different hosts and carry different plasmids. Two plasmids in B. anthracis code for toxins that cause anthrax in mammals. Bt has one that makes insect poisons. And although its plasmids seem innocuous, the main part of Bt’s genetic material codes for toxins that can cause diarrhoea, vomiting, muscle and kidney damage and liver failure. Leaving plasmids aside for a moment, this means that Bt already possesses the genes that make B. cereus an opportunistic human pathogen—one that we can normally fight off, but which can strike if we are weakened by another illness. Manufacturers insist their commercial strains are safe. “The questions you are asking [about toxins] seem geared to strains not at all used commercially,” Tracy Sorrentino of Bt manufacturer Abbott Laboratories told New Scientist. Abbott refused to make any further comment on any aspect of Bt safety. But recent research in Canada suggests that commercial strains of Bt do make B. cereus toxins. Vern Seligy and colleagues at the Canadian federal health ministry told the American Society for Microbiology in Chicago in June that, at concentrations similar to those in aerial sprays, two commercial strains of Bt killed human cells in culture, by producing toxins that behaved like those from B. cereus. “The DNA sequence information for most of the virulence genes in B. cereus is in current Bt products,” says Seligy. There are also reports of health damage with Bt. Katy Young of the Environmental Health Alliance, a campaigning group in British Columbia, says that in 1994, after Bt was sprayed to kill gypsy moths in forests near Victoria, 62 people had problems consistent with B. cereus toxins. Their symptoms included diarrhoea, vomiting and respiratory problems. Young suspects many infections are never diagnosed. While agreeing that Bt produces small amounts of cereus toxins, the US Environmental Protection Agency (EPA) says there is no valid evidence to link use of Bt insecticides with episodes of diarrhoea, and it has therefore declared the products safe. Some scientists, however, point out that commercial Bt strains could become more aggressive, perhaps by swapping regulator genes with wild B. cereus. Andrup has discovered that some strains of Bt, very similar to those used commmercially, contain plasmids that cause the bacteria to join up with other Bacillus bacteria, and pass DNA back and forth (Journal of Bacteriology, vol 181, p 3193). “This powerful conjugation system,” says Andrup, “could spawn harmful bacteria in the environment, where sprayed Bt can survive a year.” Not least, Bt could pass on the gene-swapping mechanism itself. The dangerous plasmids in B. anthracis appear unable to pass into other Bacillus species. But armed with powerful new conjugation genes from Bt, in theory, this could change. It’s an alarming scenario. However, Jackson thinks it is unlikely to happen because the bacilli rarely meet in the vegetative or “growing” state needed for plasmids to be swapped. Bt usually grows only in its insect host, B. cereus in soil, B. anthracis in mammals. This makes it extremely unlikely that Bt will ever swap dangerous DNA with its Bacillus cousins, says Jackson. Others are not reassured, however. While separate territories keep the bacteria apart in nature, they fear modern agriculture might bring the different species together. Artificially growing and spreading billions of extra Bt in sprays might cause events that are vanishingly rare in nature to occur often enough to spawn dangerous hybrids, says Andrup. “We should certainly remove the conjugation system from commercial strains,” he says. Seligy notes that some commercial Bt strains grow at mammalian body temperature and pH. Kimothy Smith of Northern Arizona University points to what are now nearly forgotten experiments in the 1940s, which showed that B. anthracis could grow in insects—where it could encounter vegetative Bt. In western Europe, anthrax has been eradicated. But recent outbreaks in Montana, Alberta and the former Soviet Union show it has not gone away. One way of calming fears about Bt insecticides would be to remove the toxin genes from the bacterium. Jo Handelsman of the University of Wisconsin in Madison is developing a seed coating made of B. cereus, which inhibits fungal infections by producing an antifungal antibiotic. To ensure that the coating is harmless to humans she is deleting the toxin genes from her strain. She says her system of gene deletion could also be used with Bt. Jackson, who uses Bt in his garden, says: “Bt should remain in use—but we must be more careful to monitor it.” Seligy goes further. “Anything to reduce the expression of harmful genes would be useful,” he says. While some multinationals, such as Novartis, have moved out of Bt production, commercial interest in using biopesticides—such as Aspergillus flavus against aflatoxin contamination of cotton—shows no signs of slowing. Eight other applications now before the EPA, awaiting approval for use in the US, concern opportunistic pathogens in humans (see examples in Diagram). If the EPA decides that these and any other bugs are “friendly” enough to be used safely,