If your chrysanthemums look stunted and ugly, take comfort. They’re infected with a parasite that may tell us a lot about how life began.

Chrysanthemums and some other plants are victims of invisible enemies with the wonderfully sci-fi name of viroids. Scientists discovered viroids in the 1960s while they were trying to figure out why potatoes sometimes grew in weird long shapes (right). They suspected that this so-called potato spindle tuber disease was caused by an infection, because the condition seemed to be spread slowly through potato fields. As for the infectious agent, it seemed likely to be a virus. After all, viruses have long been known to be able to change the shapes of plants they infect. Looking for a virus in the slow-growing potatoes would have slowed down the search immensely. But researchers figured out how to transfer the disease to fast-growing tomatoes, which became stunted after being infected.

Yet even with this advance, the scientists still struggled to figure out what was making the plants sick. They used centrifuges to get rid of particles larger than certain sizes and found that whatever was infecting the plants was amazingly small—far smaller, in fact, than any known virus. And its chemistry was different from that of any known virus. Viruses carry their genetic information in molecules of either DNA or RNA (a single-stranded counterpart to DNA). They also keep this genetic information in a protein shell. The scientists ground up infected tomato leaves and mixed them with enzymes that chop up RNA. They could no longer use that mixture to infect other plants—suggesting that the pathogens used RNA for their genetic material. And indeed, when the scientists repeated the experiment with enzymes for cutting up DNA, the material was still infectious. But then a third experiment yielded a weird result: When the scientists used enzymes to cut up proteins, the material could still make plants sick. In other words, the pathogen seemed to be nothing but a tiny snippet of naked RNA.

It took years of tests to finally confirm that the potatoes were getting sick from bits of raw RNA. Scientists dubbed them viroids (meaning “like a virus”) and went on to discover them in a number of plants, including chrysanthemums. In fact, it turned out that a disease that had nearly wiped out the American mum industry in the mid-1950s had been caused by a viroid of their own. (For some reason, no one has found an animal viroid.)

A viroid can contain as few as 250 nucleotides. That’s extraordinarily small: The human genome is 3.3 billion nucleotides long, and a single protein-coding gene may be several thousand nucleotides long. Yet that’s enough genetic information to let a viroid infect a plant and replicate itself by taking advantage of the enzymes the plant uses to replicate its DNA. Normally, the enzymes pull apart DNA’s two strands and add nucleotides to make two new strands, using the old strands as templates. But viroids trick these enzymes into copying their single-stranded RNA. The enzymes use the viroid as a template to make a new viroid progeny. Plants have evolved defenses against these viroids that prevent them from being duplicated. But viroids can often escape the attacks of their hosts, leaving the cell to infect another.

While plant scientists try to figure out how to cure crops of viroids, other biologists are fascinated by them because they strip the processes of life down so far to the bone. Recently, Rafael Sanjuán of the University of Valencia in Spain and his colleagues decided to see how much viroids mutate. Humans, chrysanthemums, and other multicellular organisms have many ways of lowering their mutation rate. They can proofread their DNA as it is copied and correct errors. Bacteria are not quite so careful, but they can still repair a lot of DNA damage. Many viruses are decidedly sloppier. What’s intriguing about this pattern is the size of the genomes involved: The higher the mutation rate, the smaller the genome.

Viroids offered a fresh opportunity to test this relationship, because they are hundreds of times smaller than viruses, the smallest genomes for which mutation rates had been estimated. Sanjuán and his colleagues infected chrysanthemums with viroids and then let them breed. The researchers then harvested the new viroids and scanned their genomes for mutations. In particular, they looked for mutations that would keep the viroids from replicating, because these must have been new. (They couldn’t be carried down from earlier generations, because they keep the viroids from replicating.) To make sure these really were lethal mutations, the scientists engineered viroids with these mutations and injected the mutant viroids into plants. The plants didn’t get sick.

Sanjuán and his colleagues found many mutations. The viroids are the fastest mutators ever found, mutating thousands of times faster than the previous record holders, they report in the 6 March issue of Science. And, as this chart* shows, viroids fall right where you’d expect along the genome size/mutation rate continuum. (The mutation rate is measured in the chances any nucleotide has of mutating per generation.)

As I wrote in my Origins essay in January, many scientists are now persuaded that the earliest form of life on Earth was based not on DNA, RNA, and proteins, but on RNA alone. Some are now building RNA-based protocells to see if they can grow, replicate, and evolve. If life did start out in this RNA world, the early protocells would have had only a few relatively short RNA molecules. As Sanjuán and his colleagues observe, viroids bear a striking resemblance to them. According to the RNA world hypothesis, the original RNA molecules of life would have had to have carried out two different kinds of functions. They’d have to store genetic information and also speed up chemical reactions the way enzymes do today. The viroids that Sanjuán and his colleagues studied contain a stretch of RNA called a hammerhead enzyme that speeds up the viroid’s duplication in an enzymelike way.

If early RNA-based life forms had genomes on par with those of viroids, they would have had a gigantic mutation rate. We could not survive with such a high mutation rate because we have such a big genome, with so many vulnerable spots where mutations could deal lethal blows. With a much smaller genome, viruses can survive at a higher mutation rate. And viroids, smaller still, afford mutations an even smaller target.

But the reverse is also true: Without a way to lower their mutation rate, viroids are trapped below a so-called error threshold. The evolution of very accurate gene replication was thus probably a crucial stage in the emergence of more complex life. If life had stayed sloppy, we’d all still be little more than viroids.