When Jesse Bloom heard in 2009 that Tamiflu, once the world’s best treatment for flu, had inexplicably lost its punch, he thought he knew why. Sitting in his lab at the California Institute of Technology, the biologist listened to a spokesperson from the World Health Organization recount the tale of the drug’s fall from grace. Introduced in 1999, the compound was the first line of defense against the various strains of flu virus that circulate around the world every year. It did not just treat symptoms; it slowed the replication of the virus in the body, and it did its job well for a time. But in 2007 strains worldwide started shrugging off the drug. Within a year Tamiflu was almost completely useless against seasonal influenza.

The WHO spokesperson explained that the sweeping resistance came about through the tiniest of changes in the flu’s genetic material. All flu viruses have a protein on their surface called neuraminidase—the “N” in such designations as H1N1—which helps the viruses to break out of one cell and infect another. Tamiflu is meant to stick to this protein and gum it up, trapping the viruses and curtailing their spread. But flu viruses can escape the drug’s attention through a single change in the gene encoding the neuraminidase protein. A mutation called H274Y subtly alters neuraminidase’s shape and prevents Tamiflu from sticking to it.

Most public health experts had assumed that flu viruses would eventually evolve resistance to Tamiflu. But no one anticipated it would happen via H274Y, a mutation first identified in 1999 and originally thought to be of little concern. Although it allows flu viruses to evade Tamiflu, it also hampers their ability to infect other cells. Based on studies in mice and ferrets, scientists concluded that the mutation was “unlikely to be of clinical consequence.” They were very wrong. The global spread of viruses bearing H274Y proved as much.

That spread “sounded alarm bells to me,” Bloom says. Something else had changed to let the virus use the mutant neuraminidase without losing the ability to spread efficiently. He soon found that certain strains of H1N1 had two other mutations that compensated for H274Y’s debilitating effects on the virus’s ability to spread from cell to cell. Neither of the pair had any effect on their own. In the lingo of biologists they were “neutral.” But viruses that carried both of them could pick up H274Y, gaining resistance to Tamiflu without losing their infectivity. Both mutations looked innocuous individually, but together they made the virus more adaptable in the face of a challenge. To put it another way, they made it better at evolving.

Such neutral mutations are also known collectively as hidden or cryptic variation. They were long ignored by most researchers, but thanks to technological advances, scientists are starting to see that they are a major driving force in evolution—including the evolution of microorganisms that make us sick. By studying cryptic variation, scientists are finding new ways of safeguarding our health and discovering fuller answers to one of evolution's most fundamental questions: Where do new adaptive traits come from? As Joshua Plotkin from the University of Pennsylvania puts it: “This is the forefront of modern evolutionary biology.”

How it works
As the flu example shows, one way that cryptic mutations can enhance adaptation is by collaborating with other mutations to produce a whole that is greater than the sum of its parts. Imagine that someone gave you a triangular metal frame or a pair of wheels. Both parts would be useless on their own, but put them together and you get a working bicycle. If you have either component, you have nothing immediately useful but you are primed to reap the benefits of the second one. In the same way cryptic variation can lay the groundwork for future adaptations.

Some cryptic mutations can also prove useful on their own, essentially keeping quiet until circumstances arise where they come in handy. And, by building up a lot of cryptic variation, organisms can increase their ability to adapt. Imagine that something goes badly wrong in your house. If you have a bunch of tools you never needed before stashed away in a cupboard, one of them might end up being good for the job or could be modified. In the same way, a storehouse of cryptic variation increases the chances that living things will be preadapted to cope with new challenges.

These ideas fit well with Darwin’s theory of natural selection, in which beneficial traits that boost an organism’s reproductive success are passed down to future generations, or “selected” to continue on. Biologists, however, are increasingly realizing that some mutations are important not because they provide immediate benefits but because they enable adaptations to occur in the future. These mutations can build up because natural selection does not remove genetic alterations that have no obvious effects on our proteins, cells or bodies.

The notion that cryptic mutations can be useful has a long history. In the 1930s Sewall Wright, one of the founding fathers of evolutionary theory, recognized that initially unimportant genetic changes could later give rise to valuable ones. Theodosius Dobzhansky, another central figure, said that species need to have “a store of concealed, potential, variability.” Even so, until very recently scientists managed to document only a few arcane examples of hidden mutations affecting the wings or hairs of flies without any proof these changes benefited the animals. “We didn’t have the tools to take it further, and the topic languished,” says Joanna Masel from the University of Arizona in Tucson. With powerful sequencing technology and mathematical models on hand, scientists have now been able to show that cryptic variation is a powerful and widespread force in evolution. In everything from flu viruses to flowers to fungi, they have found tangible case studies where useful adaptations arose from seemingly neutral mutations.

One of the clearest examples comes from Andreas Wagner at the University of Zurich and involves molecules called ribozymes, which consist of RNA (genetic material related to DNA) and function in the body as catalysts. They speed up chemical reactions involving other RNA molecules but are picky about the ones they interact with. To react with a new target, they need to alter their shapes. And to do that, they need the sequence of their building blocks to change. In test-tube studies Wagner found that ribozymes could adapt to deal with a new target six times faster if they had previously built up lots of cryptic variation. Just as in the Tamiflu-resistant flu viruses, these mutations made no difference on their own; they merely brought some of the ribozymes a step closer to achieving the changes they needed. “They had a leg up in the evolutionary process,” Wagner says.

Another example comes from studies of heat-shock proteins, which help nascent proteins to fold properly into their functional forms and also protect them from losing their function in response to various stresses, such as excess heat. In 1998 Suzanne Rutherford and Susan Lindquist from the Massachusetts Institute of Technology showed that a heat-shock protein called Hsp90 can both hide cryptic variation and unleash it, depending on circumstances.

By helping proteins to fold correctly, Hsp90 allows them to tolerate genetic mutations that might otherwise catastrophically distort their shapes. It can thus allow proteins to build up such mutations, along with neutral ones. If conditions become more challenging—such as a significant rise in temperature—the Hsp90 molecules may be in such demand that they cannot aid all the proteins that need them. Suddenly, proteins have to fold without Hsp90’s help, and all their cryptic mutations become exposed to natural selection. Some of these mutations would have beneficial effects in the challenging conditions and would thus pass to the next generation.

Rutherford and Lindquist first demonstrated what Hsp90 does in fruit flies. When they depleted the protein by exposing flies to heat or chemicals, the insects grew up with all sorts of weird features, from subtle, unimportant things like extra hairs to severe deformities like misshapen eyes. None of these changes were caused by fresh mutations but rather by existing dormant ones that had been hidden by Hsp90 and unmasked by its absence. For good reason, Lindquist has described Hsp90 as an evolutionary “capacitor,” after the devices that store electrical charge and release it when needed. It stores cryptic variation, unleashing it in demanding environments, just when it is most needed.

Hsp90 is ancient and found in plants and fungi as well as animals—signs that it is one of life’s critical molecules.One of Lindquist’s lab members, Daniel Jarosz, discovered that a fifth of all the variation in the yeast genome is concealed by Hsp90—a huge reservoir just waiting to be released. By exposing so much variation in one fell swoop, Hsp90’s behavior provides a possible answer to one of evolution’s most puzzling questions—the origin of complex combinations of traits.

“Sometimes it’s hard to envision how new forms or functions could emerge if they require multiple mutations, none of which are individually beneficial. The frequency of it should be exceedingly rare,” Jarosz says. It is a dilemma that opponents of evolutionary theory often seize on. But heat-shock proteins, and cryptic variation more broadly, provide a possible solution. When environments change, they allow organisms to make use of mutations that were sitting quietly in the wings but that in combination suddenly offer a solution to some challenge to survival. They act as evolutionary rocket fuel. “Hsp90 can help us to understand how complex traits could ever be achieved in very rapid fashion,” Jarosz says. For those in the field, it is an exciting time. “We’re really at the cusp of making big discoveries in the most fundamental question in evolutionary biology: ‘How does life bring about new things?’” Wagner says.

The disease connection
Beyond offering new insight into the underpinnings of evolution, research into cryptic mutations is suggesting new ways to look at and combat disease. It has been very hard to decipher the genetic underpinnings of many human traits or diseases, from height to schizophrenia. Even though they run strongly in families, scientists have found only a small number of genes associated with them. Plotkin wonders if cryptic variation might help to solve the puzzle of this “missing heritability.” Perhaps we should be looking for mutations that have no effect on their own but rather influence the risk of diseases in combination. “This is just wild speculation on my part, but it sounds reasonable to me,” he says.

The same thinking is being applied to other disorders. We continually provide bacteria, fungi and viruses with new challenges by attacking them with our immune system or hitting them with waves of toxic drugs. One of their chief defences is the ability to evolve resistance, and cryptic variation helps them to do this faster. Lindquist, for example, has shown that Candida albicans, the fungus responsible for thrush, needs lots of Hsp90 to evolve resistance to antifungal drugs. When she blocked Hsp90, the fungi stayed vulnerable. Cancer cells also benefit from Hsp90, because they need help in folding their wide array of unstable mutant proteins. Many scientists are now testing chemicals that block Hsp90 as potential treatments for cancer or ways of preventing fungi and bacteria from developing drug-resistance.

Others are trying to predict how cryptic variation fuels the evolution of viruses. Plotkin and Bloom are focusing on influenza. “The flu virus is evolving all the time to escape all the antibodies that it has stimulated in the human population,” Plotkin says. “This is why we have to update the vaccine every year.” Last year he analyzed the genomes of flu viruses collected over four decades. He found hundreds of pairs of mutations, where one swiftly appeared after the other. In many cases the first of the pair was neutral—it did nothing except to pave the way for the second mutation. By identifying these hidden mutations, which predate more serious ones, we could find strains that are primed for resistance and cut them off with the right vaccines. “We could, to some extent, predict the evolution of flu,” Plotkin says.

Plotkin also envisages focusing on cryptic mutations for a different end: making new molecules useful for the biotechnology industry. Many scientists are trying to artificially evolve designer proteins that will do specific tasks. Typically they look for mutations that overtly alter proteins in ways that enhance their ability to do the chosen task. But it may be useful to look for the hidden neutral mutations that could make proteins more likely to acquire useful mutations. “Understanding the role of cryptic mutations in an evolving protein could help to improve some already very useful techniques for engineering enzymes,” Plotkin says.

Applications like these are just the beginning. In many ways the study of cryptic variation has been a metaphor for itself. Knowledge and interest in the field has been building up under the surface for a long time, largely hidden from view, only to be released by the influx of new technology. “We’re really at the tip of the iceberg,” Plotkin says.