ONE night last July, biologist Anton McCaffrey was driving north towards San Francisco when a California Highway Patrol officer waved him over. "Approximate speed, 87 miles per hour. Whoops," McCaffrey recalls. But the excuse he offered to the cop was certainly original: he was having trouble concentrating on his speed because his mouse livers had stopped glowing.
Just hours earlier in his lab at Stanford University, McCaffrey explained, he had seen the first results of an experiment that could revolutionise medicine. McCaffrey and his colleagues were trying to shut down the activity of a gene in living mice, using a completely new type of drug. They had engineered the mice to produce a glowing protein whenever and wherever the gene was active. This meant that normally, these mice would be scampering around with glowing livers. But after months of painstaking research they had at last succeeded in switching off the gene and getting rid of the glow. "The officer just gave me a blank look," he says. "But he did cut me a deal on the ticket."
It was a pretty impressive achievement, even if the cop didn't realise it. No one had ever succeeded in switching off a gene in living mammals in this way before. McCaffrey's team had snuffed the gene with the help of an ancient immune system that up to a few years ago was thought to be present only in humbler organisms such as flies, worms and plants. It's a discovery that has astonished and excited biologists.
By showing the same system was lying unnoticed in mammals and could be harnessed, the work paves the way for a completely new technique for tackling human disease, by switching off genes at will. It's a tool that promises to help us attack rogue genes in cancer and beat back the viruses that cause AIDS and hepatitis.
This newly discovered immune system is called RNA interference, or RNAi, and is one of the most exciting new areas in biology. "This used to be a small, exotic field," says Thomas Tuschl of the Max Planck Institute for Biophysical Chemistry in Göttingen, Germany, whose lab first found RNAi lurking inside human cells little more than a year ago. "Now people are flooding in to explore the possibilities."
Labs and companies are scrambling to exploit its potential. Such is the hype that one British broadsheet hailed it in a lead article as "a revolution" and "a genetic discovery to change the world". And with good reason. The RNA immune system promises to give us new weapons in the war against disease, and it has certainly overturned some cherished notions about biology. But although researchers are optimistic and the results of the experiments spectacular, no one can yet answer the key question. Will the RNA immune system live up to its promise in the clinic?
Like many remarkable discoveries, scientists stumbled upon RNA interference entirely by accident. A decade ago, Richard Jorgensen, now at the University of Arizona, and Joseph Mol, working independently at the Free University in Amsterdam, were experimenting with genes for flower colour in petunias. Both of them gave the flowers an extra copy of a gene coding for a purple pigment, expecting to produce a more intense colour. But often the flowers were simply white, suggesting that the extra gene not only played dead but somehow stopped the plant's original pigment genes from working.
This discovery left the teams scratching their heads. Adding more genes should only boost the levels of protein encoded by those genes - making the flowers deeper purple, not white. Meanwhile, flowers weren't the only organisms flaunting their disregard for genetic theory. Other researchers working on the mould Neurospora crassa and the tiny soil nematode Caenorhabditis elegans were also finding that adding extra genetic DNA, or even just incomplete RNA copies could actually result in less gene activity.
The researchers were stumped. Their findings completely contradicted every tenet of textbook biology. It's supposed to work like this. The genes in a cell's chromosomes are made of a double helix composed of two strands of DNA. Each strand has a backbone that sports a string of "letters" consisting of chemical bases named A, C, G, T. The bases on the two strands pair up, A with T and G with C, zipping the strands together. When a gene is switched on, the cell "prints out" a copy of the gene's letters on a single strand of RNA, a molecule rather like DNA except with T replaced by U. This printout, called messenger RNA, then gets shuttled off to protein factories called ribosomes, which read off the sequence of letters. The RNA printout tells the ribosomes which amino acids to use to build the protein encoded by the original gene. Basically, genetic information is transcribed from DNA into RNA, which is then used to make proteins.
So according to this orthodoxy, RNA's role is rather limited and lowly - the messenger boy of the cell. Biologists took all this pretty much for granted.
Blocking the flow
But Jorgensen's peculiar petunias gave the first clues that there could be more to RNA than this simple role. Researchers realised that when they added a gene to a cell, any of the cell's own genes that had a similar sequence got shut down. It turned out that the messenger RNA from these genes was being destroyed before it could be used to make a protein. The flow of information from DNA to protein was being blocked, but no one knew how, or why.
A big breakthrough came four years ago from Andrew Fire at the Carnegie Institute of Washington in Baltimore, and a team at the University of Massachusetts. They discovered that a potent trigger for this gene shutdown was double-stranded RNA - two strings joined together just as they are in the DNA double helix. Most cells have only single-stranded RNA, but some viruses have the double-stranded variety. Suddenly the cell's motivation was perfectly clear: it thought it was under attack and was trying to close down the supposed invader's genes.
It seems that one function of this RNA defence is to attack suspicious gene sequences that might have come from viruses or other genetic parasites, rather like the way the body's main immune system attacks suspicious proteins - or policemen pull over cars going suspiciously fast. It's obvious why double-stranded RNA should set alarm bells ringing. It's less clear why some artificially inserted genes are also identified somehow as troublemakers. In petunias, one theory goes that adding the extra purple genes could, for reasons that are only poorly understood, result in the formation of strange RNA structures that trigger the defence mechanism. But however the cell sniffs out foreign DNA, once it does the cell starts shutting it down, along with any similar genes. Normally, this isn't too much of a problem because most virus genes have very different sequences to plant genes.
The discovery of RNAi was a dream come true for many researchers. Here at last was a way of shutting genes down at will. But even though the technique works well in practice, researchers are still trying to understand how RNA interference works in different animals. So far, they've had most success in animals such as fruit flies, starting with double-stranded RNA as the trigger (see Diagram). Researchers now suspect that there is a similar mechanism in mammals, except that the trigger may be different.
The pairing between the bases in DNA and RNA is key to how RNAi works. When double-stranded RNA enters the cell, its unusual structure rings alarm bells. An enzyme called DICER quickly slices it up into short double-stranded pieces, like a knife slicing up a baguette into chunks. The cell then peels the two strands apart and uses them as probes to seek out matching messenger RNA strands, which they stick to because their bases pair up exactly. Once these probes are stuck to the rogue messenger RNA, they destroy it with the help of other enzymes. This effectively shuts down the unwanted gene.
"It's much like the strategy a human programmer uses to track down computer viruses," says Fire. "They use a piece of viral sequence long enough to be specific, but small enough not to encode a dangerous piece of the virus."
At that point, geneticists had a field day using double-stranded RNA triggers to take out any genes they chose in animals such as flies and worms. But hopes of using it in human cells were scotched. That's because most mammalian cells have a completely different response to double-stranded RNA. They commit suicide in spectacular style, shutting down all protein production and putting their genetic material through the shredder. For good measure they spray out a chemical called interferon that warns surrounding cells of a potential viral invasion. It seemed to the researchers that this rather drastic defence mechanism had taken the place of the RNA immune system in our cells.
But Tuschl's team realised there was another explanation, as did Fire, working independently in collaboration with Natasha Caplen at the National Institutes of Health near Washington DC. Perhaps the system was still alive in mammalian cells, they guessed, but had simply been masked by the suicide program, which was an extra level of defence. Other researchers had noticed that you had to add big pieces of double-stranded RNA - at least 30 letters long - to trigger the self-destruct program. So the two teams wondered what would happen if they bypassed the first step of the RNAi pathway and just added ready-chopped pieces of RNA, dubbed "small interfering RNAs", or siRNAs, instead.
Back from the dead
They struck gold. Both groups found that when they fed these RNA pieces to mammalian cells, they could shut down genes at will without causing the cells to commit suicide. In other words, the ancient RNA immune system was ready and waiting in our cells, and researchers had finally found a way to bring it to life.
Mammalian geneticists were at last free to join the party, and many labs are now knocking out various genes with gusto in the hope of finding out what they do. Gregory Hannon and his team at Cold Spring Harbor Lab in New York, for example, are part of a multi-centre effort to knock out 15,000 different genes in various human cancer cell lines.
But it's the therapeutic potential of our newly discovered immune capability that is drawing the most excitement. It promises to open up a new avenue of attack against old viral enemies such as HIV and hepatitis. And in a flurry of publications this year, researchers proved that in animals at least, RNAi can produce spectacular results.
May 2002, for example, John Rossi and his team at the City of Hope National Medical Center near Los Angeles used RNA interference to reduce the activity of HIV genes in human cells by a factor of 10,000. Soon afterwards, in July, Phillip Sharp and his colleagues at MIT in Boston announced that they could slow down virtually every stage of HIV's life cycle by pummelling cells with siRNAs. Next came the key "glowing livers" experiment, performed by McCaffrey with Mark Kay and his colleagues at Stanford and Gregory Hannon and his colleagues at Cold Spring Harbor in New York state. The gene they shut down belonged to the hepatitis C virus, and it was the first time anyone had got RNAi to work in live mammals.
Anyone who feared it might be a fluke didn't have to wait long for supporting evidence. Within a few weeks, David Lewis and his colleagues at Wisconsin-based company Mirus published a similar experiment showing RNAi could suppress genes in mouse liver, kidney, spleen, lung and pancreas. In August, scientists at biotech company Intradigm in Rockville, Maryland, announced another success: they had used the method to slow the growth of mouse tumours.
Not surprisingly, companies and labs are scrambling to get on board the RNAi bandwagon. "I've been to many meetings that aren't about RNAi, but it is quickly pencilled into the schedule at the last minute," says Lewis. Even at a time when funding bodies are still shell-shocked from the dotcom collapse, this bold new idea isn't going begging for cash.
Sharp, Tuschl and David Bartel at the Massachusetts Institute of Technology and Phillip Zamore of the University of Massachusetts Medical School in Worcester are in the process of forming a company called Alnylam to develop RNAi therapeutics. Even at this early stage, they have managed to raise $15 million. Mirus has also attracted millions of dollars to pursue RNAi research.
It's not just the initial results that have impressed potential investors. RNAi isn't the first RNA-based technology to have raised hopes for a radical new treatment - but it does have key advantages over previous strategies. Only a few years ago biologists were fired up about the potential of antisense RNA - single strands of RNA that block a gene's messenger RNA by binding to it, but don't trigger the RNA interference mechanism. Meanwhile other researchers were using "catalytic RNAs", which can chop up targets on their own.
But these two technologies have stumbled because they are hard to target accurately, and act by blocking the cell's biochemistry. In contrast, RNA interference is far more precise. What's more, labs report that RNAi is a more efficient way to destroy a target, which makes perfect sense to Zamore. "With other strategies, you are trying to block some cellular process," he says. "With RNAi you are just directing the cell's own biology. You and the cell are on the same side."
That precision is good news for gene therapists. They can already add gene activity to cells, but would dearly love to be able to block genes as well. It could be invaluable in treating genetic diseases such as Huntington's, which is caused by a rogue protein that disrupts cell physiology. Researchers could use the RNA immune system to shut down the gene that codes for the rogue protein while using conventional gene therapy to add a healthy copy.
Still, despite the recent flood of papers, researchers acknowledge that RNAi therapeutics are still an unproven idea. "These are proof-of-principle experiments," says Sharp. "Between this and the clinic is a long and tortuous path." Kay agrees. "We're pretty excited about it, but the question remains if there are going to be limitations as we go forward," he says.
One obvious question mark is how easy it will be to deliver RNAi-stimulating drugs to a patient's cells - and how often it would need to be done. Both Kay's and Lewis's groups used high-pressure injection to deliver siRNAs to mouse tissues, but the effect dwindled after 3 days. Neither thinks this will work well in people because human bodies are too big. So the teams are working on other strategies. Kay's team is adapting the viruses used for gene therapy to furnish cells with genes that code for siRNAs. Mirus is using chemistry to make artificial viruses that they hope will be able to target siRNAs to particular tissues.
Delivery isn't the only issue. Controlling the behaviour of the RNA immune response once it has been unleashed could prove tricky. In plants, the RNAi response can spread from cell to cell, travelling through 30 centimetres of plant tissue. In worms, RNAi reactions can spread even more impressively: the gene shutdown can be transferred from a mother to her offspring. Whether this strange form of heredity exists in other animals is unclear, but Craig Hunter and his team at Harvard University in Boston have found that a protein involved in these properties of RNAi is also found in mammals. If RNAi can spread through human tissues, this could make it an even more powerful therapeutic - or make it harder to control.
Another limitation is that RNAi may not work in all tissues. And the precision of the technology may occasionally work against it. Since viruses such as HIV can mutate so rapidly, they may soon alter their DNA to evade a wide array of siRNA drugs. Indeed, some viruses in plants and flies have already evolved defences against RNAi, although such anti-RNAi tactics haven't yet been seen in human viruses.
What's more, it may well be important not to overwork the RNAi machinery, because it may have other jobs to do besides tackling viruses. Genomes are littered with potentially destructive pieces of DNA called transposons, which can jump from place to place, disrupting genes. In worms, for example, a failure of the RNAi machinery causes the transposons to activate, suggesting that RNAi helps suppress these renegade pieces of DNA. That raises the worry that enlisting RNAi to fight viruses for us could leave us vulnerable to our genomic parasites.
But despite these caveats, our new-found line of cellular defence is the most promising therapeutic avenue to have emerged for years. And the revolution that RNAi has triggered isn't confined to medicine: it has entirely altered the way we think about biology.
When the RNA immune system isn't busy fighting off invaders and policing parasites, it has some important civilian duties. Far from being a mere messenger boy, RNA plays a key role in controlling normal genes in a cell - a complete reversal of the traditional view of its role. It may even be crucial to the development of all animals and plants.
One the best examples comes from a gene called let-7. It was first seen in C. elegans, but researchers have now found it in many animals, including flies and humans. A mutation in the let-7 gene produces a defect in worm development. But the gene doesn't code for a protein. It makes a single-stranded 70-letter RNA molecule that then loops back on itself like a hairpin to form a double-stranded molecule.
Tuschl and Zamore showed recently that this hairpin is processed by DICER to produce "microRNAs". Rather like siRNAs, these microRNAs bind to messenger RNA, but they don't degrade it. Instead they simply stop the cell's protein factories from reading the message and making a protein. This proves that RNAi is more than a one-trick pony. It is a complex machine that can be programmed by different RNAs to perform different tasks.
When let-7 was discovered, it was one of only a few strange genes coding for hairpin RNA. But in the past year, researchers have found that a wide variety of organisms, including humans, may have hundreds of genes for microRNAs, with different tissues producing different ones. It looks like RNA plays a key role in managing the biochemistry of the cell.
In a nice twist that truly turns the biological orthodoxy on its head, it transpires that RNA can even end up in charge of DNA. In plants, for instance, the RNA immune system can trigger the complete shutdown of genes so that they don't even get as far as making messenger RNA. And experiments with yeast suggest that DICER and other parts of the RNAi machinery help to shape centromeres, the structures within chromosomes that guide them to the right places when the cell divides. If that turns out to be true for most organisms, it would put RNAi slap bang at the heart of chromosomes and biology.
For all the excitement surrounding RNAi, the remarkable fact is that its discovery didn't depend on genomics, proteomics, bioinformatics or any other high-tech innovation of modern biology. In fact, many experts have noted that nearly all the biochemical and genetic techniques behind its discovery were available 25 years earlier. As to why it took scientists so long to find this basic player in biology, there are many views.
Some experts think the dogma of molecular biology blinded researchers to the role RNA might play in influencing the flow of information. And the fact that siRNAs are so tiny also helped them elude detection. "Any RNA that small, people just assumed was degraded and threw it away," says Sharp.
However, he is convinced that the delay in recognising RNAi comes down to a simple fact: no one was looking for it very hard. "The mindset was and is that we already know everything about the biology of cells and are just filling in the details," he says. "This story should remind everyone there are still probably things out there we don't understand, or even know that they exist."