Rowbotham was unable to identify a number of the samples he collected in 1992, and he stored the mystery cultures in his lab freezer for future research. When budget cuts forced the closure of his lab in 1998, he had the presence of mind to call around and ask fellow scientists if they would be interested in any of the critters in his fridge. He had recently met Didier Raoult, he recalls, "and when I told him on the phone about the cooling tower cultures, he said he'd love those systems."
A student of Rowbotham's who had just accepted a postdoc position in Raoult's lab in Marseille took the samples south with him. Four new strains of Legionella would eventually be drawn out and identified at Raoult's lab, along with new bacteria closely related to Chlamydia, parasitic bacteria that, like Legionella, cause a variety of diseases. One last sample, however, defied all methods of examination for more than a year and a half, until La Scola turned his high-powered scope on the last Bradford holdout: Mimivirus.
In addition to its signature viral shape, Mimi exhibited what is known as an eclipse phase, a bit of telltale viral creepiness recognizable to any fan of sci-fi horror movies. When a virus penetrates a cell, it disappears inside the nucleus for four to eight hours, giving the outward appearance of complete normalcy. Then the viral particles that the cell has been coerced into making suddenly burst forth, shattering the host.
Still, it wasn't until Raoult sought the assistance of Jean-Michel Claverie, a bioinformatics specialist at the Institute of Structural Biology and Microbiology in Marseille, that the true weirdness and wonder of their monster was revealed.
At about a half-millionth of a meter across, Mimi is one of the few viruses visible under a standard light microscope. Its genome weighs in at a whopping 1.2 million letters: at least 10 times larger than a typical virus's; nearly triple the size of that of its largest viral counterpart, canary pox, in the smallpox family; and larger than the genomes of 20 or more parasitic bacteria. Moreover, within Mimi's outsize helping of genetic material, Claverie found genes for such things as the translation of proteins, DNA repair enzymes, and other types of protein. Those functions were thought to be the exclusive province of more complex cellular organisms. The boundary between viruses and complex bacteria had become officially blurred.
"We already had very large viral genomes in the database before Mimivirus," says Claverie. "But before we saw that the virus and bacteria groups could overlap, we never asked ourselves why some large viruses had, for example, 300 genes, while the typical virus only needs 10. Then we see Mimi, with over 1,000 genes, and we're thinking we have a problem with our whole concept of viruses."
Viruses come in all shapes, sizes, and degrees of sturdiness, and with all manner of strategies for getting at the cellular machinery they lack. Some batter-ram their way through the outer cell membrane. Some meld their membranes with a cell's and then suddenly revolve, like those faux bookcases in the movies, into the sacred chamber. Still others gain entry by disguising themselves as the sort of free-floating molecules that our cells routinely gobble up.
The manner of replication varies, too, depending on the virus's genetic identity. DNA viruses like smallpox, herpes, and now Mimivirus tend to be larger and more sophisticated genetically. They can exist for centuries outside a host and can afford to be more restrained when replicating inside one, making reliable, relatively error-free copies of themselves by hijacking the formula common to all life.
DNA makes a slight variant of itself known as RNA, which directs the production of the specific proteins of which all complex life-forms are composed. So-called RNA viruses are rogues: smaller, fast-replicating shape-shifters, descended from a time that evolutionary biologists refer to as the RNA world, back near the base of life's tree, before today's DNA-based organisms evolved. RNA viruses can direct the copying of their own proteins without using DNA—a shortcut that generates both more copies and more errors, or mutations. Although such activity might get you fired in the business world, in biology, mutations can offer a leg up. During unstable times—when environmental conditions shift or humans develop a successful vaccine—RNA viruses have the resiliency to adapt, outflank, and reemerge.
Influenza is the best known continuously morphing RNA virus. HIV is a particularly insidious RNA virus, known as a retrovirus because once inside the cell nucleus it reverses the DNA formula: a single strand of RNA manufactures its own double strand of viral DNA. That viral DNA is then directly spliced into the host cell's DNA and passed along with the cell's natural replication process.
There is even a newly discovered category of subviral agents known as viroids: naked snippets of RNA that lack even an outer protein coat and don't encode for anything. They are devoid of genes entirely, and yet they replicate and cause illness once inside a host. And then there are deeply derivative entities called satellites, metaviruses that can replicate only within a virus that is already busy inside a host.
Whatever neat conceptions and categorizations we develop, viruses have always found a way to poke holes in them. Scientists long assumed, for example, that viruses could only be made of DNA or RNA. Then in the late 1990s, a number of viruses were found to contain both. Retroviruses, meanwhile, were long thought to infect only animals. The only seemingly safe assumptions were that viruses will always be smaller in both physical size and genomic content than the simplest bacteria and that viruses had to have evolved after those same cellular organisms, on which their parasitism depends.
ow the discovery of Mimivirus has rendered even these two viral paradigms questionable. What Claverie calls "the final click" came after comparative analysis of Mimi's DNA with that of other organisms in life's three domains: the eukaryotes, bacteria, and archaea. Mimi, it turns out, belongs to its own distinct and extremely ancient lineage of large DNA viruses. Moreover, certain signature Mimi genes, such as those that code for the production of the soccer-ball shape of its capsid (an outer protein coat common to all viruses), have been conserved in viruses that infect organisms from all three of the domains, particularly in eukaryotes. The implications of that finding are truly radical: that Mimi, or a Mimi-like ancestor, emerged prior to the three other domains and played a key role in inventing the very cells of which humans and all complex cellular life-forms are made.
It is a difficult concept to get one's head around. Parasites, to us, are derivative, necessarily descendant from the biological entities they depend on for life. But simple does not always mean less evolved. Mimi's outsize complement of genes—so large that the virus is tantalizingly close to being an independent organism—suggest to many scientists that Mimivirus underwent reductive evolution early on and shed some of its genome, including the genes necessary to replicate on its own.
