杏吧原创

A STRANGE and disturbing force is at work in the wilds of Queensland, along the north-east coast of Australia. Venture out at dusk into the rugged hills encircling the city of Townsville, and in the failing light you鈥檒l glimpse the silhouettes of a dozen hunched creatures, each perched atop its own boulder, silent as a sentry.

Not only do these hunchbacked 鈥渞ock hoppers鈥� look mysterious, something mysterious has happened to them too. Something has caused these marsupials to undergo genetic changes at a positively alarming rate.

But what? An alien crash site oozing radiation? A covert operation to create a master marsupial? The explanation proposed by Rachel O鈥橬eill, a geneticist at the University of Connecticut in Storrs, is only marginally less strange. She claims to have good evidence that the culprit was none other than a group of mischievous viruses. What鈥檚 more, she says, the viruses wrought such profound genetic change that they gave birth to whole new species of rock hoppers 鈥� possibly in as little as a few decades.

If she鈥檚 right, Darwin is going to spin in his grave. It鈥檚 a blow for the idea that it takes millions of years of natural selection to create a new species. And evolutionary biologists may have to accept that sometimes 鈥� just sometimes 鈥� an out-of-control genome might do the same job almost overnight.

Until geneticists sat up and took notice, rock hoppers were best known for their tails 鈥� long even by wallaby standards. The seven species that live away from the Queensland coast boast a range of other novelties 鈥� from black mohawks and socks to pink patches on their heads that wash out in the rain, like logos on cheap T-shirts. By comparison, the eight species of rock hoppers that occupy the Queensland coast look pretty uniformly brown.

Like cult members, they are difficult to tell apart by looks alone 鈥� there are perhaps four people in the world who can do it. It was only when scientists systematically examined their DNA in the 1970s that they realised the creatures were separate species, and might just have something interesting to say about how new species emerge. After all, it鈥檚 not every day you get eight identical-looking species living shoulder to shoulder in the same ecological niche.

To understand O鈥橬eill鈥檚 pitch for how that might have happened, you first have to remember that it鈥檚 not just genes that vary between species 鈥� it鈥檚 the architecture of their chromosomes. The centromere, the bit that lets each pair of chromosomes link up before cell division, can change its position. The arms of each chromosome can be longer or shorter, and the DNA can be wound loosely or tightly around its protein spools, giving the chromosome its stripy appearance.

Usually, there鈥檚 not much structural difference between the chromosomes of closely related species. The llama in South America and the camel in Africa have identical-looking chromosomes, even though they鈥檝e been separated by an ocean for 30 million years. (They both have their own novel genes, however.) Similarly, a standard set of 14 chromosomes has served most marsupials since their origins 100 million years ago, and it persists almost unchanged in such far-flung relatives as wombats, bandicoots and Tasmanian devils. Even South American possums, which haven鈥檛 brushed elbows with their Australian brethren for 80 million years, have the standard set.

Macropods, such as kangaroos and wallabies, however, are a glaring exception. 鈥淭here鈥檚 a tendency among macropods to play Lego with their chromosomes, and in rock wallabies it鈥檚 just gone completely haywire,鈥� says Mark Eldridge, a marsupial geneticist at Macquarie University in Sydney, and one of O鈥橬eill鈥檚 most enthusiastic supporters.

Compare the allied rock hopper that lives around Townsville with the unadorned rock hopper, its neighbour to the south and you find some startling differences. Townsville鈥檚 finest differs from its plain cousin in having chromosomes 3 and 4 rearranged so that their centromeres are at one end rather than in the middle, while chromosomes 6 and 10 have fused together.

The chromosomes of the Mareeba rock hopper, which lives just to the north, are even more jumbled. It has two extra chromosomes fused together. In fact, judging by their chromosomes alone, the eight species of Queensland rock hoppers look as if they diverged from one another 100 million years ago, rather than 800 thousand years ago, which is what changes within their gene sequences suggest.

If this all sounds confusing, remember that the world of speciation research is characterised by incongruities. Biologists have, after all, yet to decide how to define a species. Normally, they use a mishmash of different approaches 鈥� categorising species according to their looks, such as red buttocks or webbed feet, whether or not they mate to produce fertile offspring, or according to their genetic makeup. But it鈥檚 an imperfect art: some species, including the rock hoppers, look identical, while others 鈥� 10 per cent of animals and 20 per cent of plants 鈥� can mate with different species to produce fertile offspring.

The picture of how new species arise is also murky. The theory most often evoked is pure Darwin 鈥� populations of the same species take up residence in different ecological niches, where upon natural selection gradually hones them to fit their own niche until eventually they look and act differently from one another and a new species has emerged. Lake Nyasa in southeast Africa, is supposed to have spawned hundreds of species of cichlid fish this way. Some are adapted to hiding among rocks, others to living in sandy bays, some have mouths for eating scales off the right sides of fish, others for eating scales off the left sides, and so on.

Another idea is that reproductive isolation drives speciation 鈥� a fragmenting forest splits a population, preventing interbreeding until random mutations slowly but surely create distinct species.

And then there鈥檚 the theory that might best explain those identical-looking rock wallaby species in Queensland. Perhaps chromosome shuffling is creating new species by driving a reproductive wedge between two populations. The problem is that no one 鈥� until now 鈥� has devised a good explanation for what, exactly, might set the ball rolling.

鈥淲hen you look at the 4000 species of mammals,鈥� says Stephen O鈥橞rien, head of the Laboratory of Genomic Diversity at the National Cancer Institute in Frederick, Maryland, 鈥渟ome species鈥� chromosomes are shuffled five times more than others.鈥� It鈥檚 a puzzle we don鈥檛 understand.

One group of suspects that have been kicked around for years are retroviruses, which wheedle their genetic material into our own. It鈥檚 a creepy thought, but the DNA of everything from cucumbers to cuttlefish is loaded with the genes of retroviruses that inveigled their way in millions of years ago. Once inside, they insert more copies of themselves into the host鈥檚 DNA, even turning up on other chromosomes. The copies also swap genes among themselves by a process called recombination.

But they鈥檙e sloppy operators. When a retroviral gene moves to a different chromosome, a piece of the host鈥檚 own DNA often goes with it so that you end up with, say, a piece of the long arm of chromosome 2 stuck to the short arm of chromosome 7. And it鈥檚 not just retroviruses: other types of 鈥渏umping鈥� DNA 鈥� collectively called retroelements 鈥� are also suspected of scrambling chromosomes.

Fortunately most living things have ways of silencing unwanted genes. Methylation, the simple addition of methyl groups to DNA, causes it to coil tightly around the spools of the chromosomes. In so doing, it silences retroviral genes, stopping them from playing Mr Potato Head with the chromosomes. But what if they weren鈥檛 always silenced, wondered O鈥橬eill, could that explain how the rock hopper genomes became so jumbled?

Enter Benny, the offspring of an unholy union between a tall swamp wallaby and a tubby tammar wallaby. Nobel laureate Barbara McClintock suggested 30 years ago that retroviruses might occasionally reshuffle chromosomes, and she thought the most likely place for this to happen was in the offspring of parents from two different species. They often have shuffled chromosomes, and the more distantly related the parents, the more shuffled they are.

Tammars and swampies are distant cousins, so Benny鈥檚 chromosomes are really weird. Many have centromeres that are 10 times as long as normal. In addition, part of an arm from chromosome 2 has moved to chromosome 7, while part of the X chromosome is reversed.

And when O鈥橬eill analysed Benny鈥檚 DNA she found, just as she predicted, that it was dramatically undermethylated. 鈥淚t was very extreme, and quite shocking,鈥� she says. 鈥淲e were looking for a needle in a haystack, and we accidentally just sat on it.鈥� This wasn鈥檛 the only revelation, either. When O鈥橬eill sequenced the chromosomes with long centromeres, she found they were actually made of pieces of retrovirus DNA repeated thousands and thousands of times, which she described in a landmark paper (Nature, vol 393, p 68).

Do the virus shuffle

With that evidence under their belts, O鈥橬eill and Jennifer Marshall Graves, her PhD advisor at La Trobe University in Melbourne, filled in the gaps in their theory for how chromosomal jumbles could occasionally create new species.

First, of course, members of different but closely related species must mate and produce a youngster. Then, suppose one such hybrid fails to methylate its DNA correctly right at the start when it鈥檚 just a single fertilised cell. Seizing their chance, the retroviruses awaken and make thousands of copies of themselves 鈥� thoroughly scrambling the chromosomes before the hybrid embryo reaches the size of a pea. Once that hybrid reaches maturity it just needs to find a secluded spot where it can breed with a few animals from one of the parent species. Within a few generations, its descendants will have the same chromosomes, including some of the original hybrid鈥檚 shuffled ones, and a new species is created.

鈥淲e鈥檝e always felt evolutionary change would be very, very slow,鈥� says Marshall Graves, who now splits her time between the Australian National University in Canberra and the University of Melbourne. But Benny contradicts all that, she says: 鈥淲e find major changes that probably occurred very rapidly after fertilisation. Something that we thought might take 50 million years might take 5 minutes instead.鈥�

Their idea may sound over-elaborate, but in rock hoppers at least the theory seems to be borne out by geography. The Queensland rock hoppers are crammed into a coastal strip of forests where there would be ample opportunity for the ancestral species to interbreed. By contrast, the rock hoppers scattered around the rest of the huge Australian continent probably had little opportunity for such liaisons, and their chromosomes are less shuffled and more like the standard issue.

鈥淵ou see the same thing over and over again,鈥� says O鈥橬eill. Each of the two main species of Australian grasshopper, for example, has its own unique set of chromosomes, but where their habitats overlap there are plenty of grasshoppers sporting newly shuffled chromosomes. And in northern Italy, a new strain of house mouse with jumbled chromosomes has appeared in the last 20 years 鈥� once again, at the intersection between two other species鈥� habitats.

To the uninitiated, of course, mating two different species of animals might seem an easy way to create a third. But while plants do it sometimes, it鈥檚 usually a dead end for animals, which is one of the limitations of the new theory. Although animal species can produce fertile offspring when they interbreed with closely related species, those hybrids are usually less fertile than the parent species, so their genetic contribution quickly disappears from the population.

鈥淭hese new hybrid entities probably form fairly often,鈥� says Loren Rieseberg, a geneticist at Indiana University in Bloomington. 鈥淏ut most of them cannot survive [competition] with the parental species, and so they get swallowed back up by the parental species.鈥�

To get around that problem, O鈥橬eill and Marshall Graves envisage a hybrid like Benny finding itself in an isolated spot where it only has to compete with a few other animals. That scenario would seem reasonable for rock hoppers, which dot the countryside in small groups, wherever there鈥檚 a boulder heap.

But reduced fertility is not the only reason why some researchers doubt that genome scrambling accounts for many new animal species. According to Reinald Fundele, a developmental biologist at the Max Planck Institute for Molecular Genetics in Berlin, the offspring of two different species don鈥檛 necessarily have undermethylated or scrambled chromosomes. He鈥檚 examined crosses between camels and llamas, horses and donkeys, and two mouse species, and found no evidence of either. In fact, says Fundele, the undermethylation route to scrambling chromosomes may be impossible in non-marsupials. For them, methylation is so critical to early development that the embryos probably wouldn鈥檛 survive.

O鈥橬eill, however, has a comeback. She claims you just have to look harder in non-marsupials for signs of undermethylation. Using refined techniques, she鈥檚 examined her own mouse crosses and found slight reductions in methylation that would escape the cruder detection techniques used by other researchers, such as Fundele. The results are preliminary, but these mice also have chromosomal shuffles.

What鈥檚 more, natural selection might better explain the speciation driven by different ecological niches, such as the cichlids in Lake Malawi, but her theory does a better job of explaining those species that emerge in similar niches but with dissimilar chromosomes, such as the rock hoppers, certain look-alike South American field mice, which shlep around anywhere from 10 to 42 chromosomes, and the muntjacs or Barking deer of South-East Asia.

鈥淭his is a potentially very interesting mechanism where there isn鈥檛 necessarily any natural selection driving the differences,鈥� says Andrew Hendry, an evolutionary biologist at University of Massachusetts in Amherst. Eldridge agrees. 鈥淭he rock wallabies are doing something very different from classic examples of speciation. It鈥檚 the genome that鈥檚 driving it 鈥� an out-of-control genome.鈥�

One thing is certain: the Benny effect has people buzzing. A creationist website even touts it as evidence that evolution could fit into the 10,000 years demanded by the Bible. 鈥淚 was impressed at the creativity of that argument,鈥� quips O鈥橬eill, 鈥渁nd almost embarrassed.鈥�