DESPITE its universal role in biology, evolution still poses some pretty perplexing questions. Take changes in body form. Every tree or beetle or mouse looks the way it does because thousands of genes turned on at exactly the right time and place to guide the organism from single cell to adulthood. But if body plans are the product of such intricately orchestrated programs, how can evolution ever conjure up new ones? Any slight perturbation would surely send a species tumbling from its evolutionary peak into the barren valleys beneath.
Plants and animals may have hit on an ingenious solution – bottling up evolution for times when they really need it. By squirrelling away genetic mutations, the raw material of evolution, and releasing them all at once, species may be able to leap from peak to evolutionary peak without ever having to slog through the valleys between. This happy knack increases their odds of surviving stressful conditions – nothing less than evolution on demand.
On the face of it, the idea sounds like biological heresy. Plants and animals couldn’t have that sort of control over the random process underlying evolution, could they?
Advertisement
Surprisingly, they could. Over the past few years, a handful of lab experiments have thrown up convincing evidence that organisms really can save up mutations for a rainy day. If the same thing happens in nature, then plants and animals have hit on a way to seize the throttle of evolution, accelerating it when necessary and slowing it down when not. Their storehouse of mutations may also prove to be a treasure trove of new genes for drug hunters to plunder or, equally, the time-bomb that helps explain the diseases of old age.
The lead actor in this iconoclastic drama is a so-called “chaperone” protein called hsp90. One of the most abundant proteins in animals, plants and fungi, hsp90’s job is to bind to unstable proteins and help them maintain their correct shape. In this role, hsp90 is rather like a valet, tidying up proteins that would otherwise become dishevelled by environmental insults such as high temperatures. Hence the chaperones’ other name, “heat shock proteins”.
But hsp90 is also a crucial regulator of development. As a way of silencing proteins until their services are required, cells deliberately make some proteins unstable, especially certain ones that regulate developmental pathways. One of hsp90’s jobs is to hold some of these proteins in the “standby” position. “It works on just about every [developmental] pathway you can imagine,” says Susan Lindquist, director of the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts.
The first hint that hsp90’s job gives it unusual leverage over evolution came four years ago, when Lindquist and her team mate Suzanne Rutherford were working at the University of Chicago. They noticed that fruit flies carrying a mutant copy of the hsp90 gene sometimes had offspring that looked very weird indeed. “We had eyes that grew out from the head in a stalk-like pattern, we had bristles in the wrong places, wings with different venation patterns and shapes, abdomens that were partly folded over, legs that were different shapes – virtually every structure in the adult fly was affected,” says Lindquist. The same abnormalities showed up in normal flies doped in the larval stage with geldanamycin, a drug that interferes with the action of hsp90 (Nature, vol 396, p 336).
That wasn’ too surprising, given hsp90’s pivotal role in development. But when the researchers looked more closely at their abnormal flies, they saw something much more interesting: each set of parents tended to produce offspring with a distinctive set of abnormalities, different from those in unrelated fly lineages. One might have deformed legs, another strangely positioned eyes and stunted wings. If the flies’ bizarre body plans were simply down to a shortage of hsp90 playing havoc with their developmental pathways, then the defects should have been scattered randomly throughout the whole fly population. But they weren’t.
There was something else going on, and the researchers thought they knew what it was. They suggested that the familial abnormalities were caused by “cryptic” genetic defects that had lain hidden for generations and only showed up when hsp90 stopped doing its job. That would explain why each lineage sported a different set of abnormalities: each one had its own, unique collection of hidden defects.
This makes sense given hsp90’s way of working. When all is well, the researchers proposed, hsp90 goes about its usual business of keeping unstable proteins in working order. It does this so efficiently that it can even tuck into shape proteins with minor mutations that would otherwise alter their shape and, therefore, their functions. In effect, hsp90 papers over the flaws in an organism’s genome by keeping mutations hidden harmlessly away – including mutations in the genes that regulate development. Over the generations, a lineage of flies can accumulate many minor mutations that never see the light of day.
In times of stress, though, this patina of orderliness breaks down. High temperatures, noxious chemicals, and a host of other stresses can cause an epidemic of protein misfolding. Faced with a dramatically increased workload, hsp90 can no longer keep up and, as a result, malformed proteins go unrepaired. That means stress can lead to the sudden unmasking of hidden mutations. If it happens during larval development or metamorphosis, hidden mutations in developmental genes can produce abrupt changes in shape and form.
Remarkably, then, hsp90 acts as both a capacitor for storing genetic variation and the trigger that releases it. It therefore gives evolutionary biologists a convincing molecular explanation for evolutionary change. “This looks like something that’s going to put evolutionary theory on a firmer ground in terms of mechanism,” says Massimo Pigliucci, an evolutionary ecologist at the University of Tennessee in Knoxville.
Earlier this year, Lindquist’s team announced that hsp90 performs the same trick in the thale cress, Arabidopsis thaliana – making it likely that many other organisms store variation in this way, too. Once again, drugs that interfere with hsp90 caused some seedlings to develop abnormally. “We had roots that grew up instead of down, changes in the number of root hairs, leaves that developed almost like pine needles, leaves that curled up or down, leaves that became pigmented – all sorts of things,” says Lindquist. Many of the variants looked as if they might be helpful to plants trying to adapt to new environments – more or fewer root hairs, for example, might be appropriate for different soil types and moisture conditions – though the researchers did not test this directly. Again, different inbred lines of the plants produced seedlings with different sets of abnormalities, and they showed the same characteristic abnormalities when grown at higher temperatures, even without the drug (Nature, vol 417, p 618). Hsp90 seemed to be concealing cryptic mutations in plants, too.
What’s more, once these new variants come out of the woodwork, the useful ones are likely to stick around even after the environmental stress has disappeared. When Lindquist’s team bred fruit flies while selecting for certain abnormalities, after just a few generations the flies hung onto their new shapes even when hsp90 was at full function again. Lindquist believes that selecting for these traits, which are the result of many genes working together, prompts the selected lines to accumulate more and more of the desirable gene variants, until eventually the flies exceed a threshold where the trait becomes independent of hsp90. That’s important, because it means organisms won’t lose useful mutations once the stress evaporates.
The upshot of this is that species seem to have a mechanism for delivering variation – the raw material of evolution – just at the time they need to adapt to a changing environment. “It’s almost too good to be true. Just when you need variation, it’s there,” says Charles Knight, an evolutionary physiologist at the Max Planck Institute for Chemical Ecology in Jena, Germany.
Maybe so, but Lindquist’s results actually sit quite snugly with existing theories of evolution and development. As early as the 1940s, British biologist C. H. Waddington suggested that organisms must have ways of buffering mutations that could disrupt their development. Though Waddington had some experimental evidence, many biologists remained sceptical, because no one knew how such buffering could arise. Lindquist’s work provides the first molecular explanation. “The hsp90 work came as a real surprise,” says Brian Hall, an evolutionary developmental biologist at Dalhousie University in Halifax, Nova Scotia. “Here’s this molecule we’ve known about for quite some time that could play this really fascinating role.”
Hsp90’s ability to store and release mutations also helps resolve a long-standing evolutionary puzzle – how species can make the transition from one body plan to another when intermediate forms would seem to be dangerously maladapted. By storing genetic variation and releasing it all at once, a species may be able to muster the raw material for big evolutionary leaps. These larger leaps could increase a species’ chances of finding a design that’s better adapted to its new conditions, says Lindquist. It may even have been one of the driving forces behind some of the bursts of rapid diversification found in the fossil record – the so-called “punctuated equilibrium” model of evolution popularised by the late palaeontologist Stephen Jay Gould.
For all the advantages of such a system, though, Lindquist has backed away from the suggestion – which she and Rutherford hinted at in their first paper – that this storage-and-release mechanism might have been shaped for that purpose by natural selection. “We’re not by any means saying that it evolved for the sake of evolvability,” she says now.
Most other researchers agree. “It’s actually very difficult to think up cases in which systems evolve in order to make evolution more efficient,” says Nicholas Barton, an evolutionary geneticist at the University of Edinburgh. “It’s not impossible that that sort of thing can happen, but it takes careful argument to justify it.” Instead, hsp90’s buffering ability most likely arose as an incidental by-product of its main role in protecting proteins against environmental stress.
But buffering you from genetic mutation may in the end have a less desirable side effect too: one researcher suggests that overloading your chaperone system as you get older may be one cause of the diseases of ageing (see “The shock of the old”).
While Lindquist’s experiments have focused on hsp90, it is unlikely to be the only protein playing this evolutionary game. “I think this is the tip of the iceberg. There are going to be many things that buffer genetic variation,” says Lindquist. Hsp90 is just one of a whole platoon of heat shock proteins, and other molecules may also act in similar ways. Earlier this year, for example, researchers at the University of Valencia reported that an hsp dubbed GroEL can repress harmful mutations in the bacterium E. coli (Nature, vol 417, p 398).
Though even sceptics say her experiments are impeccable and the buffering mechanism she describes is fascinating, Lindquist herself wouldn’t claim organisms necessarily enjoy these payoffs in the real world. “For actual long-term evolution to occur, first of all one of the phenotypes that’s uncovered needs to be beneficial,” says University of Chicago biologist Martin Feder. Then it has got to hang around long enough to shed its dependency on hsp and become abundant in the population. That means the organism expressing it must find a mate whose genes allow this trait to appear. “While not impossible, these events are fairly improbable,” Feder concludes.
Then, too, fruit flies and thale cress are hardly typical of most species in the wild. Besides their many generations of adaptation to life in the lab, these two species became geneticists’ favourites partly because of their unusually short life cycles, which means they accumulate mutations more quickly than other plants or animals. More sedate species might never gather enough hidden mutations for this mechanism to be important. “It’s possible that the role of heat shock proteins may be overestimated in organisms that reproduce that fast,” says Pigliucci.
Nor is it clear that providing more variation will actually prompt organisms to evolve faster. Most natural populations already express ample genetic variation to support evolution, argues Barton: “Even if heritable variation were much lower than it really is, we would be able to account for evolution. You don’t have to suppose that organisms are sitting around waiting for variability to come up.” As evidence, he notes that if you select organisms for almost any trait you choose, they’ll respond – and faster than you usually see evolution proceed in the fossil record.
These will remain open questions until someone can use the hsp system to produce useful adaptations. Feder says he and Lindquist have talked about trying this critical experiment on yeast, but their plans got pushed aside during Lindquist’s recent move from Chicago to Massachusetts.
No matter how important hsp90’s masking of mutations turns out to be in real-world evolution, Lindquist’s experiments seem certain to provide other scientists with a valuable tool. Plant breeders, for example, may be able to expose the hidden variation within a crop species as an alternative to costly and time-consuming techniques such as genetic engineering and cross-breeding with wild relatives. Developmental geneticists may be able to use a similar approach to pick apart the evolution of development. Hall, for example, plans to see whether he can uncover hidden variation in a vertebrate, the dwarf African frog. If so, he hopes to use drugs to block hsp90 at different stages of the frog’s life cycle to gauge the variation at each stage. Such a snapshot, he thinks, would reveal which developmental stages allow the most leeway for innovation and which are most conservative.
All in all, Lindquist’s results have evolutionary biologists buzzing. “This could open the floodgates to a lot of follow-up,” says Pigliucci. Not a bad yield from a bunch of deformed fruit flies.
The shock of the old
Could overworked molecular chaperones such as hsp90 contribute to cancer, heart disease, diabetes and other diseases of ageing? Perhaps, says Peter Csermely, a biochemist at Semmelweis University in Budapest. And if they do, maybe we can exploit this to cure them.
Csermely’s idea sprang from Susan Lindquist’s work on molecular chaperone hsp90, which suppresses the effect of minor genetic mutations then suddenly releases them at times of stress. Musing about how this would affect people, Csermely reasoned that over the past few generations, improvements in housing, public health and medical care should have reduced the frequency of such stresses and kept the mutations even more tightly cloaked. Since natural selection would have had less opportunity to weed out these mutations – the vast majority of which are harmful – it follows that people today ought to carry more of these hidden mutations than their forebears did. And it’s likely that our descendants will carry even more.
As people get older, though, creeping decrepitude means that their proteins accumulate more and more damage, so their chaperones have to deal with more and more problem proteins until, eventually, they can no longer deal with everything. “In an old person many of the silent mutations don’t have a chance to get repaired,” says Csermely. Preliminary studies in rats suggest this does indeed happen – Csermely’s team found that chaperone proteins in older rats were less able to cope with stress than those of younger rats. Once overload begins, the hidden mutations begin to emerge and trigger degenerative diseases such as cancer, heart disease and diabetes, Csermely believes.
Csermely’s biotech company, Biorex, is working on drugs to boost levels of chaperone proteins in stressed cells. This should delay the onset of chaperone overload – but, because the drugs don’t affect chaperones in unstressed cells, side effects appear to be minimal.
In animal tests, several of Biorex’s candidate drugs work for cardiovascular disease and the long-term complications of diabetes, and two have entered clinical trials in people. In one small trial, 20 patients with high blood pressure showed significant improvement after taking Biorex’s drug for 12 weeks. However, none of the drugs has yet been tested in a large, randomised controlled trial.