LABS involved in DNA research look much the same everywhere. William Brown’s is no exception. Located on the 4th floor of a nondescript concrete block at the University of Oxford, its benches are laden with the standard tools molecular biologists use to tinker with DNA – enzymes for dissecting it, gels for teasing those pieces apart, more enzymes to glue together the interesting snippets. But in one respect, Brown’s lab is unusual. It’s a DNA lab where nobody is especially interested in genes.
That may seem odd. After all, DNA is famous for being the stuff from which genes are made. And hunting genes, especially those implicated in diseases, is one of the surest ways for molecular biologists to attract funding, status and media attention. But Brown, his Oxford colleagues, and a handful of research groups elsewhere, have another glittering prize in sight. They want to build an artificial human chromosome so that geneticists can store and manipulate the genes they discover. Natural human chromosomes – the DNA chains along which genes are laid – are too large for this purpose. But the researchers aim to strip them down to their bare essentials and use only those bits to construct a chromosome of their own.
If it works, the payoff for genetic engineering and gene therapy could be huge. Researchers could use artificial chromosomes to insert genes into animal cells and make transgenic animals; doctors could use them to deliver genes into patients’ cells. In both cases, the new genes could be inserted permanently without disturbing natural chromosomes, something that is beyond the power of today’s genetic techniques. And even the clunkiest artificial chromosome would allow researchers to probe other mysteries, not least how 46 human chromosomes can whiz around cells without crashing or otherwise damaging each other.
Advertisement
But the researchers have a long way to go. One problem is the sheer size of human chromosomes. If DNA were as thick as spaghetti, even the smallest human chromosome would uncoil for more than 10 kilometres. Yet geneticists estimate that only about 3 per cent of this expanse contains genes, the type of DNA that codes for proteins. What the remaining 97 per cent does is a subject of fierce debate (see “Message in a genome?” 12 August), but most researchers would agree that despite appearances it can’t all be superfluous. That’s because chromosomes aren’t passive storehouses for genes. They have the power to duplicate genes and pass them down the generations. Genes may give cells the information they need to make proteins, but it is the chromosomes that give genes their immortality. The key question for Brown and likeminded researchers is how much of a chromosome’s DNA is essential to perpetuate the genes and which pieces are crucial.
Chromosomes have intrigued scientists since 1882 when the German geneticist Walther Flemming observed their microscopic motions. He found that, in dividing cells, chromosomes duplicate, pair off with their doubles at the cell’s centre and then pull away from their partner into the two daughter cells. It is a simple choreography, but executed with a precision that puts any human dancer to shame; there is about one faulty step for every hundred thousand performances.
Your average piece of DNA isn’t equipped for these elegant manoeuvres. Inject a small thread of DNA into a cell – for gene therapy, for example – and it may get left behind when the cell divides. But more often than not, it will hitch a ride somewhere on one of the lengthy cables of a natural chromosome. The aimlessness of this splicing is a major headache in gene therapy. The DNA might hop into an essential gene, for instance, disabling it and spinning the cell towards cancer or death (and a gene that infiltrates a chromosome in this way can never be removed). Just as troubling, the DNA may land in a “dead” region of the chromosome where genes languish in a dormant state. If injected DNA is never used by cells, it can’t do a patient much good.
Genetic trickery
An artificial chromosome would make such therapy less like roulette. It would duplicate itself and navigate the cell without jumping into the natural chromosomes. With some genetic trickery, the therapy could even be made reversible. It might be possible, for example, to inject an artificial chromosome into someone with a tumour and switch the whole chromosome off again once the cancer has gone. To be useful for gene therapy, an artificial chromosome would also need to be easy to produce in large quantities.
At the moment, scientists haven’t even constructed the prototype. But since 1983 researchers have known what tools they would need to build one.
That’s when Andrew Murray and Jack Szostak at the Massachusetts General Hospital in Boston purified three DNA elements of a chromosome from yeast cells. By re-assembling these, the researchers constructed the world’s first manufactured chromosome, a yeast artificial chromosome, or YAC (pronounced like yak).
Despite its fancy footwork, the YAC required only three different DNA components: an origin, a centromere and two telomeres. The origin (or origin of DNA replication) enables the chromosome to duplicate itself and all the genes it carries. The centromere is the navigation device that hooks the chromosome to protein guide wires so that it can zip around the cell without mishap. Finally, telomeres provide protective covers that stop the naturally “sticky” ends of DNA strings from hopping into another chromosome. Researchers now presume that a similar trio drives human chromosomes.
Because YACs can be made to order, they have proved indispensable to researchers who want to handle large pieces of DNA. Before YACs were developed, only small pieces of DNA, consisting of just a few thousand DNA letters, or base pairs, could be handled easily. But in YACS, molecular biologists can store hundreds of thousands – in some cases millions – of base pairs. For that reason, they have become the beast of burden for the human genome project, the worldwide attempt to sequence all human DNA.
News of the YAC inspired scientists to try to build a human artificial chromosome (HAC), or what is wryly referred to as a big mammalian artificial chromosomes (a Big MAC). “It was at that moment that many of us realised what an incredibly powerful tool [a HAC] would be,” says Brown. Even the smallest human chromosome is 25 times larger than those found in yeast, so a HAC could carry much larger amounts of DNA. That extra capacity could be important, since some human genes are huge. The gene involved in Duchenne muscular dystrophy, for example, weighs in at about 2 500 000 base pairs. And YACs have another major drawback: when injected into human cells, they behave like any other piece of DNA, inserting themselves aimlessly into passing chromosomes.
Even so, studies of YACs have produced valuable insights. In cells, each DNA component is sheathed in protein molecules which help the DNA to function properly. The first step in building a YAC was to remove the protein and separate the three DNA cogs. The DNA could then be reassembled with some customised additions: a new gene with a toggle to flip protein production on and off at will, or a “kill switch” to deactivate the chromosome after it has done its job.
HAC engineers hoped to copy the YAC approach. However, the complexity and abundance of human DNA bedevilled the search for the magic trinity of chromosome elements, says Brown. “People thrashed around doing experiments that just didn’t work,” he says.
Then, in 1988, two pieces of the puzzle came to light. Robert Moyzis and his colleagues at Los Alamos National Laboratory in the US sequenced telomere DNA from human chromosomes and found that it contained a distinct signature: a set of six DNA letters repeated over and over again. The simplicity of this cap for chromosome ends seemed too good to be true. Even if the telomere contained this DNA, said the critics, could this DNA “stutter” alone do the job of a natural chromosome end?
To find out, Brown pumped cells full of telomere DNA and waited for some to jump onto a passing chromosome. Wherever the DNA landed, it lopped off the natural end and created a new telomere. Studies by Howard Cooke’s group at the University of Edinburgh confirmed that this artificial chromosome end behaved exactly like a natural one. “After not moving for so long, we had one tool in hand,” says Cooke.
More good news was on the way from Michele Calos of Stanford University. She was scouring human DNA for origins of replication, or duplication centres. Her approach was to break up human chromosomes into pieces of DNA, inject these pieces into cells and then search for ones which could replicate unaided. But there was a problem: any DNA pieces she injected immediately infiltrated and hijacked the replication centres. Calos hit on a solution. She knew that circular pieces of DNA cannot hop into chromosomes because they have no “sticky ends”. So she concealed her pieces of DNA inside a circular DNA structure. The tactic worked. Almost immediately, Calos found pieces of DNA that could replicate under their own steam. “The trick was simply to make the DNA pieces big enough,” she recalls. “Nearly any DNA piece larger than 10 000 base pairs is able to replicate.” Origins of replication seemed to be everywhere in human DNA.
If the telomere and origin were simpler than anyone had dreamed, the centromere, the chromosome navigator, turned out to be a nightmare. Under a microscope, chromosomes look like long tubes with the centromere appearing as a pinched region, often at the centre. Scientists can guess the size of the centromere by measuring this constriction. A typical centromere in Saccharomyces cerevisiae, the yeast species from which YACs hail, is only a few hundred base pairs long while a human centromere may be as large as 3 million.
But much of the DNA in human centromeres may be dispensable. Chris Tyler-Smith has turned to nature for some hints about what to throw away. As a geneticist at Oxford, Tyler-Smith knew that human chromosomes sometimes mismanage their collection of genes, so that large pieces of DNA go missing or get deleted. In some chromosomes, centromere DNA goes AWOL and it was these chromosomes that intrigued Tyler-Smith. Those that could no longer navigate cells, he reasoned, must have lost DNA vital to their centromeres. But chromosomes that could still navigate cells must still have the essential DNA. So Tyler-Smith compared several examples of human cells where the Y chromosome had been deleted. He picked out a piece of centromere DNA that was common to all the functioning chromosomes. The good news was that this DNA is only 500 000 base pairs long – short enough to be handled with a YAC. This meant the researchers could test the properties of the DNA by returning it to human cells.
The question was whether the DNA would function like a true centromere. At first the results looked promising. The YAC carrying the DNA popped into one of cell’s own chromosomes, and afterwards this chromosome would sometimes be pulled in two directions. It seemed that the DNA had some of a centromere’s ability to control chromosome movement. But could it navigate on its own? Calos and other researchers injected cells with pieces of circular DNA carrying chunks of Tyler-Smith’s centromere. Such pieces cannot infiltrate chromosomes. Nor can they navigate cells properly. But if the additional DNA worked like a true centromere, it should have helped the circles find their way. To this day nobody has managed to get a manufactured piece of DNA to move around a human cell like a chromosome.
Why not? Crucial parts of the centromere may have been left out because they lie outside the constricted area of the chromosome. Alternatively, the centromere DNA may have become scrambled before it could be injected into cells, leaving it able to interfere with a natural centromere but unable to work alone. But the most worrying possibility is that isolated human centromeres are just too complex for cells to handle.
The signs from yeast research are not good. When centromeres from a yeast known as S. pombe are injected into cells, there is sometimes a lag of many hours before they work. This suggests that cells find it difficult to correctly attach proteins to centromere DNA. The DNA of human centromeres is more complex still. Perhaps cells find it impossible to put the whole centromere back together again. “If that’s true we’ll never be able to assemble a chromosome, like they did in yeast, by simply gluing pieces of DNA together,” says Brown.
Martial arts
Because of these complications, Brown hasn’t tried to build a chromosome from bits and pieces. Instead, he wants to streamline a working chromosome while it is still in a human cell. To do that, he has turned to a form of molecular martial arts. When a human cell is pumped full of telomere DNA, this chromosome cap delivers tiny karate chops to the natural chromosomes, breaking them and creating new functioning ends. By repeating this attack several times he can whittle away at any chromosome. He has managed to carve down the Y chromosome from 50 million to 6 million base pairs – and is still going strong.
“This is a cute trick,” says Brown. And because a centromere can be no larger than the DNA on which it sits, the known size of a working centromere (and chromosome, for that matter) gets smaller with every blow. More importantly, when the chromosome gets small enough to handle easily, say half its present size, researchers can isolate the DNA, purify it and put it back into cells. That experiment will tell scientists whether engineering a chromosome from naked DNA is possible. If this stripped-down chromosome can survive a trip outside a human cell, an engineered one should as well.
Yet, like any grand plan, the pursuit of a handmade chromosome is sure to generate important spinoffs. It may have already. Calos, of Stanford, believes her replicating loops of DNA are promising tools for gene therapy. She is testing the loops in mice – whose chromosomes are similar to humans – and they might be tried in patients in a handful of years. Even though Calos’s loops contain only one element of a fully functioning chromosome, she says they are far ahead of other available therapy methods. “You could spend your whole life building the perfect chromosome,” she says. “I realised that what we have right now is useful.”(see Diagrams)
