Salad days: the secret life of thale cress

You wouldn't look at it twice. Thale cress, or Arabidopsis thaliana, is the meekest of weeds. You need a magnifying-glass to see its feeble flowers. The leaves are minimal and confined mainly to a dusty rosette at the base of the stem. Only its narrow seed pods, arranged in pairs like a fish bone, are at all distinctive. Thale cress grows commonly on walls, car-parks and railway sidings, but is easy to ignore. You could say it's the house-fly of the plant world, completely ubiquitous and of no use to anybody.

You wouldn't look at it twice. Thale cress, or Arabidopsis thaliana, is the meekest of weeds. You need a magnifying-glass to see its feeble flowers. The leaves are minimal and confined mainly to a dusty rosette at the base of the stem. Only its narrow seed pods, arranged in pairs like a fish bone, are at all distinctive. Thale cress grows commonly on walls, car-parks and railway sidings, but is easy to ignore. You could say it's the house-fly of the plant world, completely ubiquitous and of no use to anybody.

But look again. This could be the most important plant in the world. Despite its diminutive size and general dullness, thale cress, or Arabidopsis as it is known in genetics labs around the world, has hidden attributes that make it ideal for research. First, it grows fast. The life of your average Arabidopsis is over in six weeks, but in that time each one produces a fireworks display of seeds. It's a mass-production specialist, existing only to produce the maximum amount of seed in the shortest possible time, and with the minimum of inputs. It grows as readily as mustard cress.

Second, unlike most of its relatives - cresses, cabbage, rape and broccoli - this plant has not been genetically shoved about by plant breeding. No one has tried growing Arabidopsis for food, or breeding crop varieties of it. All Arabidopsis plants, everywhere in the world - and it grows the world over - are truly wild. Its genes are nature's own, not those of breeding and artificial selection. Furthermore, the plant is self-pollinated. Inbreeding preserves hundreds of genetically different populations or "ecotypes", of which there are at least 40 in Britain alone. As with the fruit fly, there are numerous wild mutants, which can be bred in the laboratory and analysed genetically.

But the single feature that has made Arabidopsis the favourite of molecular biologists across the world is its very simplicity. Its genome - that is, the genetic life-code of the species - is relatively tiny, yet flexible enough to enable the plant to grow in hot and cold places, from mountain tops to the sea-level coasts, and from England to Japan. Even in the few places where it does not occur naturally, such as America and Australia, Arabidopsis is fast spreading as a colonist. Hence you have here an incredibly adaptable but unusually thrifty plant. Although Arabidopsis has roughly the same number of genes as a human being - about 30,000 - the genome is very "clean"; unlike most species, there's very little "rubbish" between genes. You can read it like a book, albeit a book in a coded alphabet.

The plant is, in truth, a genetic athlete, fast, efficient and adaptable. It became the unlikely choice for a worldwide project to decipher and map the first full genome of a plant in the Nineties. Knowing about this little weed allows scientists to predict things about its larger cousins. As Dr Sean May, who cultivates the plant for research, points out: "If you understand how the engine of a Mini works, it's not a big step to understand how a Ferrari engine works."

The sequence of genes, strung like pearls along spirals of DNA, took years to decode. "I've been working on this little weed for over a decade," says Dr Ottoline Leyser at York University. "It used to take 10 man-years to find just one gene. Now one person, with this new data, can do the same thing in 18 months, with a bit of luck." As a result, scientists are in a better position to modify and improve the economic value of crop plants such as soybean, cotton and maize.

When it was published, in December 2000, the decoding of the Arabidopsis genome was hailed as a landmark in biology. It involved an international effort costing £40m, in which British scientists played a key role. Understanding what makes Arabidopsis tick, says Professor Mike Bevan of the John Innes Centre, who co-ordinated the European contribution, will have profound implications for human health, as well as understanding the molecular basis of crop breeding. As strange as it may seem, for example, there are 100 genes in Arabidopsis that are closely related to human-disease genes, Bevan says.

As the first plant whose genome has been mapped, Arabidopsis is now the genetic reference plant, a template for all other plant species. It has all the genes more complicated plants need for root and seed production, genes that help mend cell walls and fight invading fungi and bacteria, and other medically useful ones, such as the genes responsible for glucosinolates, the substance that gives cress its characteristic hot taste. As one Arabidopsis scientist remarked: "Now we know what it takes to make a flower."

Scientists working on the plant can now draw on half a million basic genetic types available from Nottingham University's Arabidopsis Stock Centre (NASC). The material includes all wild forms of the plant, along with tens of thousands of cultivated forms, each of which are a mutant for a different gene. Thus there are tall and short plants, ones with white or yellow flowers, and ones that contain more or less Vitamin C than normal. Every one of these plants is available to researchers, and all are tagged with marker DNA. Once one knows the "job" of a particular gene in this weed, it can in principle be isolated and used to modify a more useful plant.

For example, the Arabidopsis gene known as "leafy" has, when transferred to orange trees, the ability to make them fruit more quickly. As Dr May puts it, "it takes away the tree's juvenility". Faster oranges mean you can speed up the breeding programmes and, in principle, obtain better quality oranges. Another current talking point is the Arabidopsis gene responsible for making Omega-3 fatty acids, a nutrient notably lacking in the average diet, but which can be magnified in this humble weed. The glucosinolates already mentioned help to detoxify a lot of the food we eat, especially well-done meat (eating mustard with beef really is a good idea). Again, these genes could be used to enhance the health-giving properties of green vegetables such as cabbage or broccoli.

If the idea of genetic engineering makes your flesh creep, there is another way of transferring marked genes, or sets of genes, from one plant to another. "GM technology," says Dr May, "appeals to the garage mechanic in every scientist. You can take a component from a Mini, modify it into a Ferrari and make a better car than either. But using a non-GM approach, you can get a Porsche and a Ferrari together to make baby cars, some of which might show small internal improvements. By selecting hidden genes using genetic markers obtained from understanding how a Mini works, you could more easily select a car with better styling, or more miles to the litre."

Plant breeding could similarly produce a super-orange using a refined method of traditional crop breeding plus genetic tracking based on the Arabidopsis genome sequence without without necessarily invoking fears of releasing potentially dangerous GM organisms into the environment.

Arabidopsis could also save lives. One Danish company, Aresa Biodetection, is even developing the use of the plant as a mine-detector. The plant often turns red in stressed conditions, such as drought or mineral deficiency. Scientists have been able to breed a GM form of Arabidopsis to change colour only in the presence of nitrogen dioxide leaking from buried land-mines. The company is using only genes already present in Arabidopsis to regulate the red pigment and, since the plant is an inbreeder, the chance of genetic contamination is minimal. But to be on the safe side, the company has also selected another naturally occurring gene for male-sterility. Field trials are set to take place either this year or next.

It is all rather ironic. Arabidopsis is not only a weed, it is not even a very good weed. "It's not persistent like a dandelion, which is difficult to remove," says Dr May. Although it occurs everywhere on suitably dry soils, it is rarely common, and hardly ever invasive. It's a weed in the slow lane, surviving if not thriving. Any other species might have become the genetic blueprint for all plants. Arabidopsis just happened to have the right qualities at the right time, just as penicillin came along just in time for the Second World War.

If ever there was a species that fits the Biblical injunction to raise the meek and humble the mighty it must be the thale cress. Look out for it next time you are waiting for the train. And when you find it, look twice.