¿ìè¶ÌÊÓƵ

Next time you catch a cold, consider this as your sinuses clog and your eyes burn: the viruses now tormenting you may be living fossils that carry molecular remnants of some of the most ancient life forms on the planet. Viruses and odd virus-like bits of genetic material that lurk within the chromosomes of most organisms, including ourselves, reproduce in a variety of unconventional ways. ¿ìè¶ÌÊÓƵs speculate that some of these processes may date back more than 3.5 billion years, beyond the origin of cells. If so, they may offer researchers a rare glimpse of the evolution of DNA-based life itself.

‘Viruses represent our evolutionary heritage,’ says Alan Weiner, a biochemist at Yale University. This startling and controversial idea is attracting increasing support among scientists, following recent discoveries that could well be the missing links in that evolutionary pathway.

Most scientists agree that the earliest forms of life almost certainly used a type of molecule known as RNA (ribonucleic acid) as their genetic material (‘The first gene on Earth’, ¿ìè¶ÌÊÓƵ, 9 November 1991). RNA is closely related to DNA (deoxyribonucleic acid), the double-helix molecules that hold the genetic code within the chromosomes of almost all modern organisms. Both RNA and DNA have sugar backbones studded with sequences of the four bases which make up the code. In DNA, two complementary strands are wound together, but RNA is a single strand that uses a slightly different sugar.

In the most primitive life forms, RNA probably reproduced by making RNA copies, at first on its own and later with the help of protein enzymes. But RNA is less stable and more vulnerable to damage than DNA, which is why modern cellular organisms have evolved to store their genetic information, or genome, in the form of DNA. At some point in the history of life, then, ‘proto-cells’ switched from making RNA copies of an RNA genome to making DNA copies of a DNA genome. And this transition from an RNA world to a DNA world – what biochemist Russell Doolittle of the University of California at San Diego calls the Great Currency Changeover – must have involved an intermediate stage in which RNA originals gave rise to DNA copies.

Beyond this vague model, however, scientists know few firm details. It all happened billions of years ago, and the actors involved were probably little more than crude bags of molecules that left no fossil record. As a result, the origin of DNA-based life is a problem rich in conjecture and poor in proof.

Genetic Oddballs

Amid this uncertainty, viruses offer a tempting modern recapitulation of the Great Currency Changeover. Some viruses, including the common cold viruses, the virus that causes poliomyelitis and some plant viruses, still use RNA as their genetic material, knocking off RNA copies as though they still lived in the primitive RNA world. Others, including the herpes and smallpox viruses, exist fully in the DNA world and reproduce using ordinary DNA polymerase enzymes. The group of greatest interest to biologists is a third one, which contains those organisms that reproduce by an intermediate method.

This group includes a variety of genetic odd-balls, ranging from retro-viruses such as HIV to several virus-like families called retrotransposons – a class of ‘jumping genes’ found within the chromosomes of conventional organisms. (Jumping genes get their name because they can change their location.) All share a single, crucial trait: an enzyme, known as reverse transcriptase, that produces a DNA copy from an RNA molecule – exactly what must have happened during the Great Currency Change-over. Moreover, reverse transcriptase is structurally similar to the RNA-to-RNA copying enzymes of some RNA viruses, which suggests an evolutionary connection between the two.

But there has been a missing link in this evolutionary chain. Reverse transcriptase and ordinary DNA polymerases all require what is known as a primer – a molecular ‘hook’ on which to hang the first building blocks for the new copy. Without this hook, which is usually a piece of pre-existing RNA or DNA, the enzymes cannot even begin their task. The copying enzymes of RNA viruses, on the other hand, simply start replicating RNA without a primer. Thus DNA-and RNA-producing enzymes appeared to work in fundamentally different ways, says biochemist Alan Lambowitz of Ohio State University, Columbus.

Late last year, however, Lambowitz and his graduate student He Wang reported what they believed to be the missing link: a reverse transcriptase enzyme that can initiate replication without a primer. They discovered it while studying a tiny, independent fragment of DNA in a strain of the bread mould Neurospora crassa collected from Mauriceville, Texas. This fragment, dubbed the ‘Mauriceville plasmid’, lives inside the mould cell’s powerhouse, the mitochondria, and reproduces itself by first running off an RNA version of its genome, then recopying this back into DNA using reverse transcriptase. The Mauriceville plasmid’s reverse transcriptase can begin its DNA copy either with or without a primer, neatly bridging the gulf between RNA-and DNA-producing enzymes.

The Mauriceville plas-mid’s reverse transcrip-tase is not the actual enzyme that pioneered the Great Currency Changeover billions of years ago. ‘Nothing in the modern world is really primitive,’ Lamb-owitz admits. It’s a question of what is closest to the ancestral form, and he says his enzyme fits that description. ‘It has exactly the characteristics you’d expect for a transitional enzyme between RNA and DNA polymerases.’

Several other features also suggest that the Mauriceville reverse transcriptase closely resembles the ancestral form of the enzyme, says Lambowitz. First, even when his enzyme does use a primer to begin copying, that primer is usually just a random bit of DNA. Most other reverse transcriptases use specialised primers that bind selectively to the site where copying is to begin. But the Maurice-ville enzyme also makes occasional use of various priming mechanisms that resemble those of several distinct classes of reverse transcriptase. These similarities to branches of the evolutionary tree suggest that the Mauriceville enzyme must resemble their common ancestor, says Lambowitz.

Secondly, the Mauriceville enzyme lacks any specialised machinery for inserting its DNA copy into the chromosomes of its host cell. The machinery for this task that is found in most other reverse transcriptases must have arisen later in the evolutionary development of the enzyme, Lambowitz reasons.

Thirdly, the Mauriceville plasmid’s reverse transcriptase resembles RNA-producing enzymes in another important way. Enzymes that copy RNA genomes – whether into DNA or simply more RNA – have to be extraordinarily choosy about which RNA they copy. Modern cells teem with all sorts of RNA molecules engaged in the business of constructing proteins. Most reverse transcriptases single out genomic RNA indirectly by looking for the primer molecule that associates with it. The Mauriceville en-zyme, on the other hand, ‘recognises’ a distinctive structure at the end of the genomic RNA itself. This same system is used by the RNA-to-RNA enzymes of some RNA viruses.

Lambowitz’s colleagues unanimously praise his underlying biochemical experiments, but most take a wait-and-see stance on his claim that the Mauriceville enzyme most closely resembles the first reverse trans-criptase. ‘It’s an interesting and provocative speculation but there are so many reverse trans-criptases out there, and they prime in such strange and wonderful ways, that it may be a little premature,’ says Jef Boeke, a molecular biologist at Johns Hopkins University in Baltimore.

Biochemist Tom Eick-bush of the University of Rochester, New York, agrees. His search for the most primitive reverse transcriptase has taken a different approach. By studying a wide variety of such enzymes and comparing their structures, amino acid by amino acid, Eickbush has constructed an evolutionary tree of the reverse transcriptases. This hierarchy offers only equivocal support to Lambowitz’s claim. ‘I don’t have any doubt he has a very primitive life form here,’ says Eickbush, ‘but his Mauriceville plasmid doesn’t look any more primitive than several other kinds of elements.’

Eickbush’s approach has its own critics, though. Reverse transcriptase is a notoriously sloppy enzyme, making thousands of times more copying errors than conventional DNA-replicating enzymes. As a result, any genetic similarities dating back to a common ancestor billions of years old would long since have disappeared in a welter of typos, argues virologist John Coffin of Tufts University in Boston. Coffin and other critics suggest that retroelements may be relative newcomers in the parade of life. In their view, the predecessors of reverse transcriptase may not be the ancient enzymes of the RNA world, but ordinary, modern DNA polymerase. If they are correct, the Mauriceville plasmid’s simple, unspecialised enzyme would be the product of degenerate evolution from more complex reverse transcriptases.

¿ìè¶ÌÊÓƵs simply do not have enough data to say whose views are correct. Everyone involved in the debate agrees that it will likely remain in the realm of uncertainty for some time – perhaps forever. ‘What you’re trying to do is trace out something that’s untraceable,’ says Eickbush. ‘It is ultimately an intuitive argument,’ says Lambowitz. ‘It’s not a proof.’